✌️2-min PMM: Sendspark's Demo Wall, DigitalOcean's Media Agency, Prosci on Competitive Intel, symplr on GTM.
February 2024, Part 1 release
📈 4 MICRO [PRODUCT MARKETING] CASE STUDIES
[1] Use the 95-5 rule to integrate lead gen and brand advertising to transform your go-to-market (GTM) strategy for long-term profits.
The CMO of symplr suggests you break your buyers into 'buying now' and 'looking to buy later' groups. Per the 95-5 rule, lead generation captures 5% of in-market buyers, while brand advertising creates demand among the 95% of out-market buyers - crucial for long-term profits.
[2] Create a demo wall for new features to reduce the noise from your free plans and identify your ideal users faster.
An early startup can attract a ton of free users through built-in growth loops without ever turning them into the right customers they want. Sendspark chose to add a demo wall to a new feature to identify users willing to pay. This manual step helped clean out their clogged self-service pipeline.
[3] Build a robust competitive intelligence (CI) culture by identifying role models, adding CI to onboarding, and making it easy to share intel.
McKinsey states that you need incentives, systems and skills, and role models to change how people do things. Using that idea, Prosci proposes you spotlight key folks within your company who contribute actively to CI, add CI to the new hire onboarding process, and turn intel sharing into an effortless process.
[4] Launch an external writers' program to scale 'how-to' content aimed at technical users like developers.
DigitalOcean built an in-house media agency complete with a head of content, an editorial team, SEO gurus, and a design and production team. Additionally, they launched a writers' program to meet their goal of creating tutorials targeting long-tail keywords.
📚 1 BOOK & TOP 3 INSIGHTS
“Reimagined: Building Products with Generative AI” by Shyvee Shi, Caitlin Cai, Dr. Yiwen Rong
[1] You can use the cooking analogy to figure out what developing an AI model looks like. You need the "right ingredients (data), a recipe (model architecture and hyperparameters), culinary skills (organizational capabilities), state-of-the-art kitchen appliances (GPUs), and the final garnish (your AI 'killer app' - ex: chatbot, recommendation engine)."
[2] The VP of AI at Cisco (Barak Turovsky) identifies 3 factors to review to identify the problems best suited for Gen AI: "(i) Accuracy: How crucial is the accuracy of the information in your use case? (ii) Fluency: How important is it for the generated content to read naturally? (iii) Stakes: What are the risks if the AI-generated information is incorrect?"
[3] On deciding whether to use a pre-trained LLM through API or develop an in-house model: (i) Do you want to validate a hypothesis quickly? Aim for the plug-and-play benefit of GPT-4 or BERT. (ii) Is the market for your product proven? If yes, go for in-house to get complete control of the data. (iii) Are you concerned about speed to market or building a highly specialized product? (iv) Does your industry stipulate regulations on data storage and processing? (v) Do you have access to the necessary talent for maintaining an in-house model? (vi) Is the product a long-term play?
🧠 5 CURATED MARKETING THINK PIECES
[1] If AI Were Conscious, How Would We Know?
[2] Growth. What are you and where do you belong?
[3] Is LTV to CAC the Nickelback of SaaS Metrics?
[4] The Future of Prosumer: The Rise of “AI Native” Workflows
[5] Why AI Will Bring the “Tinder Moment” to VC