Search Results

Starbucks ran a BOGO deal disguised as referral campaign Instead of the usual buy 1 get one offer, they encouraged people to bring a friend along. I want to discuss the untapped potential of Starbucks' disguised referral program. Getting a friend means bringing anyone — your boss, parents or even your ex. This gives people more possibilities to invite people. Instead of running a traditional, boring BOGO scheme, Starbucks turned it into a referral marketing campaign at scale. There were a few advantages New user activation Resurrection of a churned user Building habit loops for existing users and driving up loyalty Maximizing reach through influencer partnerships And Starbucks promoted the offer via influencer marketing, because distribution is everything. Did Starbucks Miss Out on Retaining New Customers? But I wasn't sure if they'd retain these newly acquired users. It wasn't clear from their campaign. So I went to Starbucks on the 7th of May to figure out what they'd do to retain new customers and drive up repeat purchases. Surprisingly, they were doing nothing. 🤔 We went in. Paid cash. Had the coffee. And we left. First impressions matter: Starbucks' overcrowding problem It was so crowded and people had to queue up outside the store in order to be let in. This was clearly not the kind of experience what Starbucks stands for — and why people pay for. So, it'd be harder to retain first-time visitors who have a sub-optimal experience. Missed opportunities to close the referral loop On the surface, I couldn't find any initiatives that were meant to close the referral loop. So I went to look more closely and I found some. 1. There were coupons printed out There was a pamphlet that could be used as a discount code. There was a time limit to redeem to get people into the habit. It looks like bi-monthly is the natural frequency that Starbucks is aiming to nurture users into. But it wasn't placed in a way where customers could see it. It was next to the cashier but was facing the baristas. The baristas didn't even give it to me when I placed the order, nor when I collected my order. The reason why this coupon is important is so that they'd nurture habit building and bring users the second time in the store. 2. Starbucks Card Starbucks also has a card system that requires you to load ₹ 200 before making your first purchase. In our case, it was possible to get our ₹ 400 order for ₹ 190 if we had loaded the card beforehand. However, we were rushed through the queue and didn't notice this option, nor did the baristas mention it. If we had used the card, Starbucks would have had our email or phone number to try and win us back if they noticed we were not returning. Additionally, we would have had ₹ 10 left in our account to spend. Oh and unfortunately, one of the offers on the page was no longer valid, so it was wasting our attention without providing anything in return, adding to our cognitive load. Conclusion In my opinion, Starbucks did a great job. One key takeaway for companies running buy 1 get 1 offers is to consider running them as give 1 get 1 offers. Additionally, it's important to actively promote good offers rather than relying on users to discover them on their own. Starbucks relied on influencer marketing to do so. Yes, they missed some opportunities to keep me engaged, like promoting the 25% discount more prominently or incentivizing new users to invite more friends. They could also streamline the Starbucks card creation process to prevent baristas from discouraging customers. Despite these issues, I'm certain Starbucks will consider this a successful campaign and will continue to refine it to generate even more revenue in the future. Thanks for reading my take! I write them to add to my swipe file and to remember it when the time comes.

Index Good landing pages examples My test with Framer Framer CMS tutorial (text-based) Framer CMS tutorial (video-based) SEO landing pages are better than blogs Veed has these landing pages that drive more traffic than all of its blog posts combined. See the traffic on SEMRush. 168k monthly visitors just from a handful of pages. SmallPDF also has similar landing pages. And each of these pages bring in millions of visits each month. Look at /edit-pdf driving 1.8M monthly visits! These landing pages are fairly simple to create compared compared to blog posts. I was able to create 20 of these in a few hours. Veed has a simple CTA that directs people to the main product. Veed also uses deeplinks to improve the UX and ranking. SmallPDF has an embed right within the landing page that people can interact with. My successful test with Framer I launched a quick and dirty test and let it simmer. It was something I did without a developer or a designer. Even with hardly any effort or any real optimization, you can see that it's growing in the right direction. It's also bringing paid conversions. A Framer tutorial for marketers Framer is a real competitive advantage so I created a little tutorial guide below if you wanted to check it out. I re-created Veed's landing page in Framer but you'd just have to copy/paste your Figma components in Framer. It's easier than you think. Then, create a simple landing page. It's a drag and drop editor, just like Figma. You can duplicate and create as many as landing pages as you want. I recommend creating 20 variations to test with and understand which performs. After you know what's working, you can convert a single landing page into a CMS template. A Framer CMS template lets you edit the content but the structure remains the same. Traditional CMS platforms make it easier to write blogs but to create landing pages, you need to go to a dev. Framer changes this status quo. You can duplicate pages in Framer's CMS just as easily. Framer CMS Pricing: On their basic tier ($15), you can create one CMS template, and on their premium tier ($25), you can create ten. In my opinion, Framer should have allowed everyone to use CMS for free, even if they prompted an upgrade pop-up when publishing the site. Restricting people from using the feature makes it harder to discover. However, I do not have enough knowledge about their product to comment. It could also be that the CMS is still underdeveloped and leads to churn if people see it. 🤷‍♀️ Scaling it up even further If you wish to scale it up to hundreds of posts, you can upload a CSV file to Framer, and it will auto-populate the data. You can use either ChatGPT (instruct it to output as a table) or Google Spreadsheet's OpenAI integration. Video tutorial for Framer landing pages and CMS Learn to create SEO landing pages with Framer in 5 minutes. Thanks for reading! Best, Khushi

“Learn the rules like a pro, so you can break them like an artist.” ― Pablo Picasso The status quo for SaaS products is onboarding. But most onboarding funnels add more friction than they seek to remove. And if activation is your goal, there's only so many steps or friction that you can remove. At Streamline, we provide icons and illustrations. People can browse, customize, and export as many assets as they want. And it's not just the web app that is ungated. Our Figma and Lucid plugins as well as our desktop app is ungated or has open funnels. There's literally no barrier to using the product. Humble beginnings.. A indie-maker created a zip file with some icons. It went viral. People asked for more. And the indie maker turned into a full-blown company. After I joined, Streamline moved from a transactional based business model to a subscription based model. People had to sign up to use the free product back then. Streamline's current onboarding.. Check out our onboarding: 1. You land on the icons homepage with all free sets at the top 2. You click into a set 3. You customize colors and size before exporting in the format you want. 4. And that's it. No watermarks. No low-resolution images. No signup needed. No account. No credit card. No questionnaire. Note: We do ask users to attribute to us if they use for free. The free tier is generous enough for most users and acts us our growth loop to generate high quality backlinks. The good stuff of open funnels.. People experience the product without friction and are delighted because it delivers on what they were hoping to get. One big benefit for us is via programmatic SEO. The more time people spend on the page, Google rewards us and ranks us higher. Activation explodes because there are no onboarding steps that lead to drop offs. Everyone skipped the line. More product data, perhaps 10x more than what'd you'd see with a traditional gated experience. This helps us understand which searches have no relevant results and if some icon sets convert better than the others. More users to run experiments. Instead of relying simply on homepage A/B testing, we now have more users in the product. The not-so-good stuff of frictionless onboarding.. Tracking is extremely hard. You have to stitch pre-login and post-login data together. If your audience is tech-savvy, you have to invest in server-side tracking to bypass adblockers that block client-side. Around 40-60% of devs, designers and tech-savvy people have ad blockers turned on. Users once lost are lost forever. If you fail to activate a user, you can't hope to win them back via onboarding emails. Lifecycle campaigns won't cover up for an average product experience. You're setting up users to fail. Some users require more hand-holding or your product has complex features that you can't just let people explore on their own. No sunk costs to keep users invested. It's easy to quit and make another search on google. Conversion rates will be lower because more people explore the product. In a traditional setup, these people would've dropped off already in an onboarding. No segmentation data. Onboarding usually helps to get information around industry, company size, role, traffic source, etc. And without one, your cohort analysis capabilities are limited. No user data means harder to qualify users or pipe them into sales funnels. You can't even customize the emails based on what their role / jbtd is about. You can customize only based on what you've seen them use. To get people to sign up, you'd have give away even more value. We layered on a free trial to get people to sign up. At the end of the day, you have good friction and bad friction. You have to choose whether to keep both or remove both. Thanks for reading :) Khushi Lunkad

Hi, I'm Khushi, and I lead the growth and marketing efforts at Streamline. Streamline is an icon and illustration library with a generous free tier. In fact, our free tier is even more generous than the paid tiers of some of our competitors! The story of Streamline Streamline's founder released an open-source UI icon set consisting of over a thousand icons. This set achieved tremendous success and eventually led him to achieve financial independence. Lots of backlinks According to SEMRush, Streamline has nearly 100k backlinks, and Ahrefs rates our domain with a score of 76. Open-source products typically ask users to provide a backlink when they use the product. Open-source strategy For any successful business, two key elements are necessary: an audience and a revenue strategy. By creating a (good!) open-source product, you can effectively address the first part of the equation. It's important to note that "free" and "open-source" are different concepts. Freemium models offer some free features but place restrictions, while open-source provides everything possible and more. Solving revenue Generating revenue becomes more challenging when users' needs are already met with a free product. Moreover, Streamline is not the only open-source product available in the market. Companies like Google and IBM are also competing for the same market share. The history and the future We observe an increasing number of companies developing a defensible strategy based on open-source. If the paid product focuses on managed services while the free product is self-hosted, an open-source offering effectively discourages newcomers from entering the market. Examples : Red Hat: For a long time, Red Hat sold services for open-source software and scaled it with human resources. Hadoop: The paid tier had more intellectual property than the open-source product, but scalability was challenging due to imperfect user experience. between the two products. Streamline: Streamline represents a third-generation open-source software. It builds differentiated intellectual property around the open-source core and offers differentiated open-core software as a service. (I use "open-core" because that's the most commonly used term to refer to Streamline). People pay for more than just support, as they gain access to the Figma plugin, Lucid plugin, a desktop app, and ten times more icons than the free tier provides. Distribution-first, not product-first Although this viewpoint may be controversial, I believe more founders should consider a distribution-first strategy rather than a product-first strategy. At the very least, this approach ensures that your company won't fizzle out in the early stages. Take Netflix as an example—its biggest growth driver is the fact that it comes pre-installed in set-top boxes. By solving the distribution aspect, you can buy time and gather more user feedback, which is essential for success! Thank you for reading! Disclaimer: Opinions, insights and everything is pretty much mine. So, take it with a grain of salt!

Strongly advocating to set up a feedback collection system is perhaps one of the most impactful things I've recently done. It allows people provide feedback without being bothered by a chain of replies from customer support. Today, we have a very simple form which pre-populates as much information as possible (email, plan, Mixpanel ID, etc). This is how the form looks like for logged in users... For users without an email addresses, we ask them to share it. It's not a mandatory question to have fewer drop offs and more feedback. The Case Against Feedback Historically, our team received feedback via customer support tickets, tweets to our brand account, or replies to our marketing emails. So, it's not like there's zero feedback coming in! What this does is that there's no trigger moment that can push your team to add a feedback tool. Without a trigger moment, motivating someone to take action becomes more challenging. Inherent Culture and Beliefs If your company doesn't have a simple feedback collection system in place, the feedback loops take far longer to cycle through. As someone working in growth, shorter feedback cycles is crucial! It's our bread and butter. If you don't have a feedback tool, it's likely because there's of an inherent culture to not ask for feedback for reasons such as these: So, it's hard to make a case for adding a feedback button and using tools like UserSnap that demand a certain level of development support. But...I absolutely wanted a feedback collection tool I had my own reasons for this... Feedback from multiple channels can be delayed. If it's not easy enough, people will just not do it. Earlier, when people hit an error page, our ask to them was publicly tweet at us. 😅 I'd rather connect our designers to our users so they don't have to take my word for why a certain design decision or a product decision can hurt us. It's a way the entire company becomes more customer-centric and learn together instead of having varying levels of understanding about our customers. I was unofficially proclaimed as the best bug hunter at Streamline. Sometimes I can be a bit nitpicky too so having a large enough audience do that at scale gives me more mental peace. Feedback is a gift and you can never have enough of it! Marketing was being asked to add a feedback collection email in the lifecycle email campaigns, which isn't the ideal place to ask for feedback IMO. So, how did we deploy a feedback collection system? One thing I've learned is that you have to let people make decisions for themselves. Your job is to provide them with information and share the context. Then encourage them to react to trigger moments. And prime them into getting to that decision. It's pretty much like sales. Selling isn’t something we do to people. We do it with them. People buy products; we don't sell. Similarly, I can't make decisions for people with higher authority. I do it with them. The tricky thing about feedback is that stakeholders need to experience it to know its value. And they can't experience it unless they implement a solution. Our Trigger Moment: Streamline was launching a redesign and rebrand! It was a huge change — new logo, new branding, new UX, new monetization plans. A massive amount of team work and months of work! This could have been a significant trigger moment but it was heightened by the fact that we had to ship earlier than planned to meet a certain deadline. So, we had to rely on our users quite a bit to tell us incase we missed out something major. But even still, it was challenging to install a software like UserSnap because it required dev resources at a time where we were constrained. For us, we have tools like Tag Manager that make integrating software too easy. But for product people and devs, they are usually more wary of the integration costs. There's more management, more setup, more time, reliability issues etc. So we ended up embedding tally.so which is an alternative to Google Forms or Typeform. Now Tally isn't great as a feedback tool because it doesn't get us console data or any other data that is needed for better bug reporting but it gets us started on the right track. It was easy to implement and I bet we will graduate us to a more sophisticated tool in the future. Did implementing Tally save us time? Maybe not - it probably took us the same amount of time as it would have to implement UserSnap. Final Thoughts Stakeholders were surprised how often feedback came in. Interestingly, I was surprised that it didn't come in as frequently as I thought it would. So, I guess everyone either under-estimated or over-estimated! 😅 It saved us a lot of time and revenue. For example, we found out that our pricing pages had a mismatch. A simple typo led to lost revenue. We shipped new features that broke people's workflow. And all this feedback came pouring in. Feedback is very costly. People spend time away from your product writing feedback so make sure you don't ask too much of it! Our team was very swift in getting user feedback into the product. We shipped things in days, not weeks! Thanks for reading! Best, Khushi

A while back, I had to run A/B tests to improve the activation rate. If you've never run in-product A/B tests before, it can be a little overwhelming so here's a simple guide on how to get started. The first test The first test we conducted showed a significant lift, increasing our core action from 14% to 21%. It was a tiny change, but it made a huge impact. And got buy-in to run more experiments. The second test After we working on activation, we went a bit deeper making sure each part of the funnel was tested. We found a user insight that we wanted to actively solve for. The result? If you look at the data below, you can see that the percentage of people that successfully went and selected a plan increased by 28%. And all other health metrics improved. The test concluded that if site traffic remained constant, we had improved trial starts by 6%. I also want to acknowledge that it was a relatively big change (involved positioning and messaging) that required a week of design and dev support. When we first launched the test variant, the test failed. So, we watched some Hotjar recordings. Discovered a few bugs and some design flaws. Everyone jumped in to help and we quickly fixed it. Had we lost faith in the test early on, we wouldn't have done anything to improve it. You can start via these 3 ways. I've over-simplified them so you know where to look and what to look for. Option 1 Step 1) Use a feature flagging tool like Flagr to conduct A/B tests. Developers will create two different environments (control and test). Step 2) Then send that data to Mixpanel, linked with the users' device IDs. Step 3) Next, analyze the data in Mixpanel. Mixpanel will directly calculate statistical significance for you, which is helpful. Option 2 Step 1) Send data from Segment directly to a tool like Statsig . Or via code. I chose Statsig because it was used by companies like Notion, Microsoft, Flipkart. It was self-serve (no need to speak to a sales rep to get started). They also offer a nice startup program if you wanted that? Statsig really speeds up the process of running experiments. You get to see all your metrics easily and can filter users with the user profile data that works best for you. I haven't completely explored the product myself, so it's a good idea to do some research on your own. In my own journey, I've evaluated Eppo, Optimizely, Taplytics, Split.io, and Launchdarkly. I finally chose Statsig. Step 2: Conduct experiments in Statsig. The process is similar: developers set up an experience, and you run a test, sending data to Statsig. Step 3: Then make a decision Option 3 Step 1) Utilize the experimentation tool in your analytics platform, like Mixpanel Experiments . Step 2) Ask developers to code the experience and run the test. What to test? To collect qualitative data for this test, I sent an email to users who signed up but did not activate in the product, asking for clear reasons why they didn't. It worked well but only after a bit of tweaking (adjust copy, timing etc). Here's the final template that worked well (h/t to Hillary for sharing): Subject : Brutal feedback for {company name} Body : Hi {first name}, I noticed that you signed up for {company name} but never did {core action}. Any chances you'd share why?—even a single sentence would go a long way in helping us improve the platform. Thanks! {founder's name} Founder at {company name} Start here with the user research if you want. Then validate with data. And finally, move on to building your test using the three options I shared above! Thanks for reading! Best, Khushi

Everyone wants to build with AI but it's much harder to find an article from the perspective of an AI researcher, who experiences it from the other side. So, I invited Joy to write this article. He's also my brother :) Joy's experience spans both large-scale initiatives and resourceful projects on tight budgets. He's done a few cool things: Worked with the Government of India's Meteorological as well as the Agricultural department to build models for them. Was awarded a research grant of $2.8M from Google to invest in his AI research. He's also worked with me at Streamline. If you're looking to build with AI for your SaaS/tech org, reach out to him on Linkedin or via this email joylunkad@gmail.com . This blog is written to help anyone who wants to build something using AI but cannot figure out where exactly. It helps you build a strong intuition about what AI can do, especially for you, so you know how to find problems that can be solved and have a preliminary idea about how expensive it might be to build it. So let's dive right into it. What can AI do? Generally, I want you to think it can AT LEAST do whatever humans can do, given ENOUGH data and compute. Any process that has the following structure and satisfies the following constraints has an extremely high chance of being solvable by AI - Input → Some Intelligent Processing → Output Where, 1. There is readily available input and output data, or it can be procured. 2. The input contains all the necessary information for a human to generate an accurate output. There is no need to include the person's skilled prior knowledge or general knowledge in the input. And, If prior knowledge is needed to solve the problem, then you should be able to formulate it as another solvable problem, ie, follows that structure and also satisfies the constraints. Are there any sharp bits that I should keep in mind? These being the only constraints just goes to show how many problems in the world are just waiting to be solved. But just because a problem is solvable, doesn’t mean it is feasible to build a solution. Gathering enough data and compute can be extremely expensive. In general, the lesser the prior knowledge that needs to be built in, the cheaper and easier it gets. And if a problem requires a ton of prior knowledge but someone else has already incorporated it into another model, the costs to train an AI become significantly cheaper. More data and more compute almost always translate into better results and similarly, if insufficient, the AI will be brittle, and make unpredictable and hilariously stupid mistakes. You can also try breaking down a big unsolvable problem into multiple small solvable problems, chaining their inputs and outputs in such a way that their ends look like the ends of the original big problem. You can do this in cases where you can't find enough data for the big problem. The benefit of solving a big problem without breaking it apart is improved performance on the final task. This is usually a very big boost in performance comparatively. The benefit of chaining smaller problems when you could solve a big problem is that it usually significantly reduces the expenses. What do I do after identifying a problem that fits in that structure and satisfies the constraints? I recommend that you have a simple discussion about it with an AI expert. There are many datasets and models out there, either open-sourced or behind an API, that can be used to bypass the expensive steps. This talk will help you get a crude idea about how feasible it actually is, and more often than not, it would be much cheaper and faster to build than you originally imagined and at the same time, more novel/revolutionary. I could go into detail about patterns and nuances to look out for that might help you identify and differentiate between ideas that are genuinely hard/expensive and low-hanging fruits but that might make this a very long and exhausting read with diminishing returns, and even then I wouldn’t be able to make an exhaustive list. You now know everything you need to know to identify solvable problems in AI. The blog ends here but if you want to develop some pattern recognition and clear up any misconceptions, I wrote examples of some problems being solved by AI. I highly recommend going through them. Example 1: Art Generation (Eg: Midjourney) I am using this as an example as it's not exactly obvious how it fits into the structure and satisfies the constraints. So, this example shows how far those rules can be stretched, helping you find opportunities where it looks like there aren’t any. If we rewrite the problem, we might get a clue. How about “creating realistic images and art from a description”? This allows us to put the problem into the structure → 1st constraint - For the input & output data, we could use the billions of images scattered across the internet, and we could use their captions as their descriptions. 2nd constraint - For an artist, a text-based description is enough to create a related piece of art. We can now be sure that it is possible to build such an AI tool. These captions are not always the description of the image, and also a single image can be described in many ways. Thus, our data is not perfect. Also, our model needs to learn a lot of prior knowledge about people, animals, objects, etc as well as some understanding about the world. We now know that these issues will make it extremely expensive to build it. Example 2: Automatically Detecting Fractures from X-Rays This is a pretty simple example, but it might clear up any misconceptions. Prior knowledge is only required if it is completely absent from the input and output data completely. If it is implicitly present, ie, it can be learnt from the relationship between the input and the output data, then we don't need additional steps. X-rays (input) → A doctor examining the X-ray → Report (output) 1st constraint - Hospitals should have huge and high-quality databases containing X-ray images and their reports. 2nd constraint - X-rays, maybe MRIs in really hard cases, are all a doctor needs. As for prior knowledge, the doctor only needs the knowledge of the skeletal structure which is available to the AI in the input data. Now that we know it is possible to solve it, and it will be relatively easy. Example 3: Self-Driving Cars Vision from the car's cameras (input) → A driver → some driving decision (output) 1st constraint - For the input, we require extensive sensory data from car cameras. As for the output, correct driving decisions can be collected from the steering wheel, accelerator, brake, etc. This data doesn't exist naturally, necessitating the preemptive installation of cameras and sensors on cars. Additionally, we must persuade customers to purchase these equipped cars to gather the necessary data. This poses a significant challenge, but it is achievable. 2nd constraint - For a human, it's just decent motor skills and some practice, hence it is self-contained, ie, just cameras are enough. Now that we know it is possible to solve it. Let's figure out how expensive it might be. This looks easy, but a sharp bit here is that it is very difficult for AI bots to do things in the real world. The input data must be all-encompassing, capturing every aspect of driving, from road conditions and weather effects to the unpredictability of human behaviour. The greater the amount of implicit prior knowledge hidden within the input data, the more extensive and varied the data required. Although the problem remains solvable, it's harder than it looks. Example 4: Facial Recognition for Security Current and verified images of the target’s face (input) → A person checks if it's the same person → Approves Entry (output) 1st constraint - Companies like Google, Meta, and Apple can ask users to tag themselves and others in images. Using this, you can create pairs of current and verified images. 2nd constraint - Self-contained, and no prior knowledge is needed. Now that we know it is possible to solve it, and it will be relatively easy. Thanks for reading! Joy Lunkad Linkedin | Email: joylunkad@gmail.com

// This post has strong opinions that are loosely held. I also acknowledge that I'm taking inspiration from other industries which might not be accurate. But it's a starting point to solve an industry-wide problem we face. I'm open to feedback and suggestions. Everyone knows that marketing and growth teams are the first to be laid off. These teams have the shortest tenure in tech. Often times, they get fired because of one of these reasons: "The company has to pivot." "We have a retention problem." "Marketing can't acquire customers." "Growth can't grow revenue anymore." Not all layoffs are bad. Some are valid but there's a reason why marketers/growth people have the shortest tenures compared to other departments like product management, engineering, and customer success. I was inspired to find a solution and asked a few questions to myself: Are growth/marketers fluffy? Can they really not sell? Are they all words and no results? And assuming all of that is true, then Why is this the case? And what can we do about it? This investigation led me to explore whether the problem lies elsewhere? First, please walk with me through some examples. Then, I'll stitch everything together and share a hypothesis. #1 Intuit's most successful vs. a total failure ad Intuit is one of the most successful SaaS companies in the world with a valuation of $180 billion. Their earliest product was an accounting software. Campaign #1 To run a direct advertising campaign in PC Magazine, Intuit had to bet its entire net worth of $125,000. If it failed, the company would be bankrupt again (they had almost gone under twice before, and at one point, they had to stop paying employees' salaries). The founders wrote this direct response ad themselves after consulting with a marketing friend. Take a look at their old ad copy from 1986: The result? Huge success. The copy was drafted by Intuit's co-founder who had previously worked at Procter & Gamble. If you've got a few extra minutes, watch the interview below starting at the 33:42 mark. Campaign #2 Eventually, Intuit hired a marketing agency. This agency created a two-page ad featuring a bald lady who had seemingly pulled out all her hair doing accounting. Do you think it would perform well? Source: In$ide Intuit Book The ad was expensive to produce, with all the agency fees. Worse yet, it brought in a grand total of 4 (four!) customers of the millions of people that saw it. Yikes. No wonder the founders were upset. I think tech is relatively easy to sell. There's a clear value-exchange and risks are offset by a subscription business. "I think SaaS is beautiful because it's a pure value exchange between customer and company. And as long as the customer is receiving what they think is enough value, they'll continue to pay for your product. So it creates this really great incentivization structure for businesses to solve real problems." — Dave Burson, Head of Product, Growth & Monetisation at Canva Intuit built one of the best products. They became the market leader in 2 months. They had done lots of market research. If you watch the interview above, you'll see they mention 'research' in almost every 3rd sentence. Founder was ex-P&G and that marketing background is incredible. But who is to blame for the failed Intuit ad? The agency, the company, or just the process? To answer that question, let's take a step back and see if companies outside of tech have solved this problem. Life's not so easy in other industries. #2 FMGC companies approach differently At bigger companies like Unilever or P&G, Marketing == Product team. Marketing decides what to build, for whom to build, why to build it, how much to price it, and how to sell it. Sure, sometimes we end up with products that are over-optimized for marketing. Look at any supermarket aisle. But the products sell. They aren't built by one team and then simply handed off to the marketing team to "sell." Revenue and product accountability lies with one team. It's why PMs at Airbnb are accountable for revenue. In tech, product marketing often ends up with the task of "hey, we've built this, now let's get customers." This approach doesn't effectively address demand. #3 Baskin Robbin's 31% Discount Consider another example: Baskin Robbins, the ice cream company, has maintained a consistent message since its founding in 1947. Their logo has always featured the number 31. This number wasn't arbitrarily chosen; it signifies the "31 flavors" concept, offering a different flavor for each day of the month. If a user is well-retained, Baskin Robbins asks them to deepen usage by ordering more of the same flavor (get a tub at a 31% discount). The 31% discount is a clever trigger moment and a time-lapsed deal which, once expired, would require users to wait for about 60 more days. It associates Baskin Robbins with the 31st of each month. Upon closer inspection, Baskin Robbins has effectively implemented: Usage and Frequency Mapping:  Something typically managed by Growth teams today. Trigger Effects:  Leveraging the 31st of each month. Product:  Introducing a new flavor to attract a new niche. Activation:  Encouraging the trial of multiple flavors. You might think this entire strategy [ read case study ] was conceived by someone at Baskin Robbins. However, the idea for "31" actually came from the Carson-Roberts advertising agency (which later became Ogilvy & Mather), in 1953. So, the difference is clear And it makes sense why. Marketers don't understand the possibilities that tech can build. That's why in tech we have product managers. People with the ability to do market research and communicate to devs. So, you have sales, marketing, and growth teams all working on leads. The product team is good at identifying a pain point and making the best possible product in the world. But then, product managers might not have revenue accountability. If it doesn't sell well, the short-term solution is always to market more. Here are some examples of what is written in my college marketing book by Philip Kotler: Growth is essential for the success of any firm. Thus, to be a long-term market leader is the goal of any marketer. Marketers can try to increase the amount, level, or frequency of consumption. (What Reforge Growth courses talk about) In targeting new customers, the firm should not lose sight of existing customers. (Poor line expansion in SaaS). Ironically, some food firms such as Hershey's have developed smaller packaging sizes that have actually increased sales volume. (Pricing and packaging). While trying to expand total market share, the dominant firm must actively defend its current business. Boeing against Airbus. Staples against Office Depot. And Google against Yahoo! and Microsoft. How can the leader do so? The most constructive response is continuous innovation . (Marketing does product management). Six types of defense strategies when you're a dominant firm: position, flank, preemptive, counteroffensive, mobile, contraction (PM work). Market challenger strategies: attact the dominant firm (high risk, high payoff), attack similar sized firms that are not doing the job well, attack regional firms, attack status quo ( lemonade insurance ). Market follower strategies.. and so on.. What you'll notice is that marketing does everything except manage the supply chain. So, it decides what to build, when to build, what to price it at, and where to sell. It does NOT decide how to build it. That resides with the R&D teams and manufacturing plants. Going back to our Intuit case where the bald lady ad failed...maybe the problem isn't with marketing or with the company. It's with the process. The reason why the agency failed is because they didn't quite get it. And they didn't help shape the product. They didn't understand the nuances. And potentially, they might not have been allowed to do so. Who knows? One of the agencies I worked didn't do any research, forget shape the product. They just interviewed clients and delivered work. Things would obviously not perform. I would spend extra time outside work on the weekends to get user research done. Taylor does a good job summarizing the idea: My recommendations If we want to solve this in tech, I've got some recommendations. These are just starting points: 1) Involve growth/marketing early on. Maybe they're called the "Growth" team. But they should be a part of the process much earlier on. It's hard to market a product you don't love or have faith in. 2) Growth/Marketing should understand tech opportunities and limitations in a better way At GitHub, all their marketers know how to code. At Streamline, I took one of the most intensive design courses to speak the same language as our designers. But, I did not put the same effort into learning about frontend/backend terminologies when that would've probably been a more meta skill to develop. 3) Let marketers/growth teams write their own briefs When recruiting agencies, see if they're writing their own creative brief. Mediocre ones will do as you ask them to. Top-tier ones will write their own. A friend of mine shared Ogilvy's campaign brief if you wanted to see how it's done. #4 Ogilvy's Client Brief An example brief for Fanta. Just look at some of the questions. What do they need to know?   Product facts are not usually decisive when people choose between brands, but if they are, list them here, or attach an addendum to the brief. What do they need to feel?   Emotive campaigns usually sell harder. What emotional response are we looking for? (Avoid clichés like “friendly,” “trusted,” etc.) Why should they care?   Brands that address issues bigger than their own category earn more time and attention. Why should people talk about our brand? Marketing has always been doing growth All this growth stuff around A/B testing, in-product tests, statistical significance has been used by marketing teams since ages. Even when there was no tech. Marketing gets a bad rap because, well most marketing is bad. So, I'd say if you worked in marketing/sales/growth or any of these ancillary functions and were laid off, it could because of one or more of these reasons You didn't truly influence the product. You haven't considered the full impact of the marketing lifecycle (awareness, activation, monetization, retention). Or maybe your work wasn't good enough — that's ok! We live and learn :) I'm happy to see that the industry is moving away from spammy marketing and looking at user journeys holistically. If you liked this article, you should read up " The Top 5 Reasons CMOs Get Fired " by Carilu Dietrich who's taken Atlassian public and is an advisor to hypergrowth companies. Thanks for reading! Best, Khushi

[This article is a WIP. I plan to sponsor newsletters and list down what I learn as I go] Elena Verna The first one I sponsored lately was Elena Verna's newsletter. I learned about PLG from Elena, took most of her courses in Reforge, and followed her to her Growth Scoop newsletter. Writing is actually fun and engaging. Hers alongside Kyle Poyar's Growth Unhinged newsletters are the only ones I read often. I think she and Jonathan on her team truly set the bar on what sponsorships could look like. They truly went above and beyond There were a few things I liked: First, they were able to match our topic with the topic of the newsletter. Second, we even got a featured image with the "Get your free churn reduction analysis". Third, the post linked to another blog post of ours in the article itself. They even honored a second CTA towards the end of the article. My ask was this, Khushi: Can we also have a link at the bottom of the newsletter? Sometimes, I'm so curious about the topic Elena writes about that I want to finish reading it, before I click on an ad. If it's towards the end, there's no other CTA that competes with it for attention. And people won't need to scroll back up. If this list wasn't already enough, they even helped write a blurb for us and gave me feedback on the graphic when it didn't look quite good on the first try. My takeaway is to sponsor posts you already love when you're just starting out as a brand. Then, you can increase your brand awareness budget. I wrote them this message: This partnership is as good as it can get! I could not have asked for anything better. Thank you <3 Our cost was $3000 for this post. It's publicly available here to sponsor . For the second time around, there wasn't enough time to discuss a few items so I had to opt out of sponsoring her newsletter. And Elena just offered it in for free. 🫨 I was playing carrom with my family, celebrating my dad's birthday. And that email came through. It had Churnkey's name at the top. I went through so many emotions all at once. First was an adrenaline rush. I was thinking, "Wow, what an article. Glad we're featured." Then, a little bit of regret followed. "Was I stupid to opt out? This is an incredible article." Then came worry. "Had I messed up the comms and they didn't see we opted out?" Then came relief. "No worries, I'll convince the team to pay and honor it." And it took me 30 minutes to type an email. Before I was done overthinking and rewriting the email again and again, I heard back from her. It said, "it's on the house". What! My jaw just dropped. It wasn't an accident. We were intentionally featured a second time around. This second article of hers was even more popular on social so I definitely would've regretted opting out. Like, a lot. But yeah, it was quite classy and kind of her to do this for us. If you'd like more metrics (click rate etc), I'm happy to share some more of those if you'd like to contact me . Otherwise, you can sponsor her newsletter . There's a lot that you get: a newsletter, a social post, and if you're lucky, maybe that post would be tied in to your topic. The only downside is that the shelf life for a post is around 2 weeks, after which it is behind a paywall. But, I'm clearly biased. I love her content and would do absolutely ANYTHING to be a part of the content she puts out. So, yeah. So far, so good. [Next Post Sponsorship Will Be Updated Here]

The cheapest way to learn in SaaS is probably to buy the product and see what they're doing. Dropbox Revenue From a numbers standpoint, Dropbox has 700M users, of which are 18.16M are paying  customers (around 2.5% of all users). 2.5% is a high enough number for a PLG company. They grew revenue by 7% (~180M) in 2023. The first quarter of 2024 saw slower growth in terms of revenue though. To learn from Dropbox, all I needed to do was pay $30 and see everything I could as without having to join the team. What a steal. I gathered 130+ screenshots across their entire app, and labelled a lot of stuff 🥵 The company I work for has a paid plan with Dropbox so I know that creatives are their most important target audience. Creatives have a few challenges. They work with very heavy files and must keep them organized. They also work with clients and have an approval process. Every product that Dropbox built was about virality. Choosing creatives as an ICP also has virality baked in if you think about it. More on that to come in a different blog post. Homepage The homepage is focused on conversions. You see ‘Find your Plan” at the top. The secondary CTA is sign up for free. The fifth fold prioritizes work use case over personal in the visual hierarchy. Forces people in the Systems 1 thinking by giving two options instead of just one. The footer also focuses on the "Get your plan" CTA Idea💡 Direct people to the pricing page from the homepage where they can get a free trial Activation The first modal window I see in the onboarding asks me to upload a file and share it. When I try to share it, I see a paywall. This is brilliant! First, the feature is optimized for shareability. Dropbox stands to gain from this. Second, they show me a nice feature (password to a file). The share button is disabled. The feature is highlighted in blue to draw attention. If I didn't want to pay I could click on the "Create and copy link" text, and it would still work. Idea💡 The first feature you introduce to can lead them to a shareable feature. When you introduce a paywall, you might want to let people opt-in for a trial. Buying a plan When I click on "Buy essentials", I'm not taken to the pricing page but rather led to their plan directly. Yearly is selected by default for obvious reasons. The one odd thing is that I can't start a trial. It was available on the public facing pricing page. I signed up with multiple accounts. So, here's another thing that Dropbox is testing. Dropbox will also ask you in the onboarding itself which plan would you like to choose. You can see that the free plan is almost hidden below the fold. They also offer to remind me when I have a few days left before my trial (but I didn't really read that copy). Different layouts for the same plan I installed the desktop app and haven't upgraded yet. On the final screen of the onboarding, it tries to sell me into the "Plus" plan. Again, it doesn't mention I could start a free trial on the CTA. Anyway, the imagery is great! It compares the free vs the paid plan. Although, 2 GB on the free plan is 0.1% the size of the paid plan limit. Not 33% that this visual demonstrates. Tip 💡 Compare the free plan vs paid plan visually. Might need an art director. Now, this is interesting. You saw earlier with the Essentials plan, I never saw Dropbox Plus. But now I do. Plus comes with no free trial, since it's a cheaper plan. I also got a little confused with their naming. How was Essentials more expensive than Plus? All prices uses 1 and 9 so it got even more confusing. There are a host of clever things Dropbox does Showing how close am I to the plan limit (Using 39KB of 2TB) with a progress bar No free trial vs free trial Show both monthly and yearly plans at once, and calculated in both frequencies Look at the differences, even in the features shown on the side. Dropbox lets you bypass their free trial. It's a good anchor to the free trial. You can request an invoice too. Manages expectation that it needs to be a larger order (team of 15+) They offer multiple payment providers (although some may not work well). Dropbox doesn't let me abandon the purchase flow. Once I clicked on the upgrade option via the onboarding flow in the desktop app, it doesn't let me simply abandon it. Lifecycle Emails Dropbox emails are very inspiring. They have the perfect email previews, use lots of liquid and HTML, is super on-brand, ties every email to their core action metric, and use proper sub-domains for sending so it never hurts deliverability. There's a lot to like! Email 1 Email 2 Upsells Dropbox does a great job at upselling me into a plan. In the UI, you'll see "Invite Team Members" plastered at nearly three different places. If you were to click on it, you'd see this modal that asks you to upgrade into Dropbox Business. So, essentially I was upsold while I was in the middle of a trial. And I can't purchase just one more seat, so I need to purchase a minimum of three seats. Once you upgrade to business use cases, the invite flow is also incredible. Your team members are assigned tasks by you. Two are selected by default. One of the tasks is to upload their personal folder. The more team members upload their personal folder, the more likely will you hit those storage limits if any. Another cool thing Dropbox does, it it will ask all your team members to invite more team members in the onboarding itself. They customize the reasons for inviting differently for team members vs admin. Admin might care about security but team members might care about other things. I'll talk more about their invite flow in another blog post. It's extensive! If I were to get a trial on the business tier, it wouldn't give the bare minimum of 3 licenses but rather give me five. Feature Discoverability A big reason why people pay is how deeply you are integrated with their workflow. Many users might churn stating that they left because a feature wasn't available. When, you probably had that feature and they couldn't discover it. The beautiful thing about software is that people pay when they receive value and stop paying when they don't. Feature discoverability plays a strong role in making the first part of it happen. The coolest thing that Dropbox does is that each product has their own tutorial/onboarding. Dropbox's Replay had the best onboarding of all times. My mind was blown. You can watch it here . Can't embed it unfortunately, no idea why not. They also added a PDF file because one of the features was to "Send and Track" files. This PDF would come in handy as I discovered another feature of theirs, hidden under their most used button. My first thought when I saw this was, "Huh!? You can edit PDFs these days? Wow". This is probably so useful. The thing about feature discoverability is that people might not even know what they need until you show it to them. And Dropbox introduced me to 10s of features in a matter of a few minutes without being intrusive. Such as this one: "you can convert file formats with Dropbox!?" Sometimes, it would trigger the good old modal window when I opened up a new product. Other times, it would highlight the button enticing you to click on it. When I uploaded a file, it automatically opens up in the sidebar and the first thing I saw was this share button. And it doesn't stop there. Dropdown gives you reasons why you might want to share. In-product onboarding is more natural. They open up a feature I haven't yet explored. If they detect that I have explored a feature, it won't open up. This button on the main dashboard keeps changing based on which feature I might have not explored yet. When to monetize They monetize the number of templates, and it is not just 1-2 but a healthy limit of 15. Perhaps, it triggers a sales conversation. A niche little feature that lets you compare versions, is an add-on. Btw, the upsell you see in a feature disappears after the first click. So, they don't continuously spam the user. And also open up room for few features. Contact sales is not visible on the cheapest tier. I saw a popup only when I hit the business tier. Failed Payments When you login, this is the modal window it’ll show to you. You see this banner with a countdown at all times. Plus a "Add billing details" as a CTA that replaces the upgrade button. Thanks for reading, I'm going to improve this article in the next couple of months. There's a lot to learn!

Zombie users are paying subscribers who pay you but don't use your product. They get counted as active subscribers, when in reality they should be considered as churned subscribers. How to measure zombie users in your product? Step 1 - Identify your engagement metrics This could be anything like pageviews, sending a newsletter (eg: Mailchimp), creating a draft (Jasper), sending a message (Slack), or creating a board (Mixpanel). Step 2 - Create a filtered view of your users I like to keep my annual and monthly subscribers separated but you can cluster them together. Here's a rough idea of how that could look like. Then sort the report by value. Customizing your subscription health 1. Using % trends You can add an event with all active subscribers. Then use a formula to find the percentage of healthy active users or zombie users (eg: zombies / total subscribers%) 2. Segmenting further You can also create three groups: healthy, at-risk, and zombie. So, you can try to win-back at-risk users. This is important to do because Zombie users might come back later and request for a refund. They might also not be happy with the brand. What should you do about zombie users? I really like what Slack does with Zombie users . As soon as they track that a user has not been active in the past 30 days, they don't bill the workspace owner. Why exactly is Slack doing and isn't doing? 1. Always charge for 1 user even if the user becomes inactive. This is probably because Slack has a cost to serve in terms of data storage costs for the workspace. 2. Don't refund the charge but instead add a credit They aren't refunding for the month. Instead, they're saving a little bit of cashflow plus ensuring that the customer's bank statements remain clean. I don't like using a lot of payment apps like Google Pay because they offer cashbacks. Cashbacks ruin my account statements. Why is this a good billing strategy for zombie users? People often hesitate to start a subscription because they're worried they will forget to cancel it. This pricing decision reduces friction to conversion. It's also good for brand and word of mouth. Clearly, I'm writing about them :) And it's also good for the customer. They aren't shocked or feel like an idiot when they see their statements and realize they forgot to cancel. But I acknowledge that this pricing decision can come from a place of privilege. Some companies can't do it if they need the cash. One of the companies I worked for signed an annual contract with Sendbird (a messaging service). Now, development time took longer but Sendbird wouldn't extend or cancel the contract. This was a tiny startup paying thousands of dollars for a service they weren't using. We wrote to them a few times but they just stopped responding. In a world of Sendbirds, be a Slack :) Thanks for reading! Khushi

A long list of little things I discovered in the activation flow (part 1). If you only have time to read a few, I recommend 4, 9, 14, 16, 20, 26, 30 Potential opportunities? Not sure if this is an A/B test or a bug, but I did not see this screen in the onboarding process. After the segmentation questions, it took me straight to the modal window (?pathway=share). To reproduce, you'll need to sign up via the direct google identity login system via a work email. I also see "personal" recommended when I sign up via a work email address. It only happens when I use the email/password combo to sign up. 2. Cookie banner could be a little more subtle. I saw the cookie banner and the Google identity login on the same screen. Almost ate up half of the screen. 3. Users need to scroll to see the next question. (I'm on my smaller laptop so shouldn't affect most users). It could lead to higher incomplete rates. Suggestion would be to auto-scroll and align to the next one. Progress bar also doesn't progress as I answer all four questions since they're on the same vertical flow, essentially making it feel like the progress is static. Another suggestion would be to have a tooltip over Continue . Or some action when I hit Continue . When you hit back, your progress is lost. It would be nice to partially save progress similar to how Typeform/Tally do. 4. Perhaps this entire area that says Drop files here to upload could be clickable and not just the Upload button. A good example is YouTube Studio, though not perfect. 5. I lack business context here but would it be better to let users upload both files and folders at once? Why should someone need to make an extra click and bifurcate their search? 6. Image previews only appear after a few hours. After I upload the file, it would be nice to see the preview appear. 7. The start sharing button took me to a place where I could take no action, at least at first glance. The Share file button was disabled. A suggestion would be to group the "Create and copy link" button with the "Share file" button. It seemed a bit odd, even though I could eventually figure it out. 8. 8a. You probably know this, but the purchase flow opens in a new tab and not as a modal window. Not really an issue but thought I'd share incase you see an opportunity. 8b. It also doesn't say "Start Free Trial" unlike what the pricing page states. 9. Even after connecting my Google account, it doesn't pre-fill contact data. I'm thinking that was the purpose? Typing in, doesn't bring up any contacts. 10. There are a few links that open automatically btw. The Android page is one of them. Here's a bug report . 11. Great visual in the desktop app but it doesn't scale correctly.  2GB is 0.1% of 2000GB. The visual makes it seems like it's a third of it? 12. Essentials is more expensive than Plus. Plus is cheaper than Essentials. Was a bit confusing from a naming POV btw. 13. Probably need a different name for the app and the folder. Harder to re-open it because it accidentally crashed. 14. Entered card. Payment fails. I see no reason or error message. I just keep clicking on the button and nothing happens. 15. It asks for my PAN number. PAN number is a very sensitive number related to personal income taxes. I have never seen another company ask for it. My 2c is that if PayTm is asking for all those details from Bill, they are doing you dirty. When I pay via UPI, I should only be asked for my UPI ID (similar to a Paypal ID). It's one simple field. PayTm is likely creating an account in their software if they ask for my pan card details. PayTm's also not doing very well and could resort to doing such things to inflate their sign up numbers :( 14. There isn't anything like address number. It's a duplicate of zip code/pin code in the Dropbox flow. You could remove it. VAT is also deprecated, and replaced by GST. 15. Payment via UPI goes through and I see it in my bank account. But Dropbox never sees the light of it in the UI. In fact, I could see I had two subscriptions active and yet, it won't remove the banner in the dashboard. I'll update this if a charge goes through in 30 days. 16. Emails trigger far too often. Same emails are received twice. Both variants of an A/B test are sent a couple of minutes apart. Security new sign in email is received even I set a password/verify my email for the first time. 17. Kind of odd to not have a tier for 2 users. 18. Upsell (essential to business) required me to re-enter my card. Address was pre-filled though. 19. Doesn't auto-pull contacts from my work email address. 20. Desktop app for Dropbox is wonky. Kept crashing for an entire day. I couldn't reopen it. Reason was weird. Couldn't even open this message properly. 21. Still says "essentials" at the top when I have upgraded to "business" 22. Even when I see the "Billing Method Updated Successfully", it refuses to remove the banner. 23. Share works on one folder at time. Maybe have it work on multiple folders? 24. Surprised I never saw that you have this feature in the UI. So cool! 25. Replay onboarding asks users to upload a video and she waits for a few seconds for us to upload. I don't see any button in the same screen. 26. Upgrade bug in replay. I click on Learn more Takes me here. 27. Tiny, but that "Feedback" link isn't easy to see. 28. Could reduce the spacing at the top. 29. Have you considered allowing people to embed Replays in their website? It doesn't seem like it works today. 30. All CTAs in app don't mention that free trial is available btw. It would improve the CTR in the button copy if it was clearer that there's a trial available?