There is another play I like even more when you have strong users. Give your champion users extra credits to build something useful, then get them talking about what they shipped. If they are active on Reddit, they can recommend your product in a thread where someone is already asking for a solution to the problem they solved. The developer gets credits to keep building. You get a recommendation from someone who has actually used the product. That is a pretty good trade. All of these recommendations are even more relevant today, thanks to the rise of GEO. ## **2. Build nurture around product behavior** One of the best exercises in a PLG motion is to look at what users do between signup, activation, and payment, and build nurture around the behaviors that actually move them forward. Most developer products already have this data. You know when someone creates a project, deploys an application, creates an API key, uses a feature, invites teammates, or reaches a usage threshold. **The question is which of those behaviors actually predict conversion.** Vercel is a good example. Its core activation path is straightforward: a developer connects a Git provider, imports a repository, and gets their first deployment live. Vercel describes the first deploy as taking about 30 seconds. That gives the marketing/product team a useful activation signal. If someone has connected their repository but has not deployed, the right nurture can help them complete that step. If they are already deploying regularly and moving toward team usage or plan limits, the communication can reflect that stage instead. When I spoke with Hank Taylor about this, he said: > The best campaign trigger isn't a content calendar, it's product telemetry. When someone is approaching their credit cap or just crossed an activation threshold, that's the moment a short personal note lands, because it's about their situation, not your roadmap. > > — Hank Taylor, Marketing Advsior, Cognition, Browserbase, G2i Inc. Start with the users who converted and work backwards. Look at what they did before activation and what they did before paying. Then compare that with users who signed up but never activated, or activated but never paid. Find the behaviors that consistently show up before conversion and make those your nurture triggers. If you want to improve signup-to-activation and activation-to-paid conversion, this is one of the first exercises I would run. ## **3. Original technical content is becoming more valuable** There is a lot of pressure to produce more content with AI, but the useful content is getting harder to fake. Google continues to emphasize unique, valuable content for both traditional search and its AI features, while its spam policies target scaled content that exists primarily to manipulate rankings.We're excited to announce VectorAI DB, the first vector database purpose-built for high-performance, reliable AI at the edge.
— Emma K McGrattan (@emmakmcgrattan) April 28, 2026
RAG isn't dead. It just can't run in the environments that need it most.
In manufacturing, 46% of AI pilots never leave the OT network. Healthcare,… pic.twitter.com/l2pErARhGi
For technical products, this creates a very specific opportunity: publish things that contain knowledge your competitors cannot generate by rewriting the same ten search results. That means original benchmarks, technical experiments, implementation guides, architecture decisions, production lessons, developer workflows, and research from your own product. Here are a few examples of what that looks like: - **Comparisons:** [**Doppler’s workload identity vs. secrets**](https://www.doppler.com/blog/workload-identity-vs-secrets) and [**ZenRows’ Apify alternative**](https://www.zenrows.com/blog/best-apify-alternative-for-large-scale-scraping). - **Technical benchmarks:** [**Actian’s analysis of AI infrastructure and cloud repatriation**](https://www.actian.com/blog/databases/what-37signals-cloud-repatriation-taught-us-about-ai-infrastructure/) and [**ZenRows’ web scraping API benchmark**](https://www.zenrows.com/blog/best-web-scraping-apis-benchmarked). - **Market research:** [**Actian’s look at vector databases in 2026**](https://dev.to/actiandev/whats-changing-in-vector-databases-in-2026-3pbo). - **Technical tutorials:** [**OpenRouter’s LangChain tutorial**](https://openrouter.ai/blog/tutorials/langchain-chatopenrouter-setup/) and [**Actian’s HIPAA-compliant AI infrastructure guide**](https://www.actian.com/blog/developer/how-to-build-a-hipaa-compliant-ai-ecosystem-without-the-cloud/). - **Technical opinions:** [**Doppler’s OIDC edge functions piece**](https://www.doppler.com/blog/oidc-edge-functions) and [**Actian’s analysis of vector database pricing**](https://www.actian.com/blog/databases/the-hidden-cost-of-vector-database-pricing-models/). You can use AI to make the process faster. You can use it to structure research, create briefs, analyze data, and improve distribution. But the expertise, evidence, and experience still need to come from somewhere, and that somewhere is your engineer. ## **4. Find relevant conversations on Reddit** Postiz has grown to more than **$2M ARR**, and Reddit has been one of its distribution channels from the early days. The play is simple here: find conversations where people are already discussing the problem you solve and become part of those conversations. I would use [**Octolens**](https://octolens.com/) or [**Brand24**](https://brand24.com/) to monitor Reddit for competitor names, alternatives, product categories, and phrases that show buying intent. For a web scraping product, I would track terms such as “Apify alternative,” “best scraping API,” “web scraping API,” and “blocked by Cloudflare.” When a relevant thread appears, I would have someone from the team join the discussion and answer the actual question. I would not try to sell. If the product has a limitation, I would say it. If another product is better for a specific use case, I would say that too. Developers appreciate knowing where a tool fits and where it does not. I have seen Redditors respond much better to that kind of honesty than to someone claiming their product is the best at everything. The goal is to become a useful person in the discussion. The product can come up naturally when it is relevant. ## 5. Make transparency part of your marketing Transparency has become a norm in developer marketing. The interesting question is what you are willing to make public. Here are the areas I would look at: - **Your marketing numbers.** PostHog publishes its marketing budget and breaks down where the money goes across ads, events, sponsorships, and other channels. - **Your product decisions.** Publish why you built something, changed it, or decided not to build it. GitLab makes a large part of its product planning and roadmap public. - **Your failures.** Cloudflare has built a strong reputation around detailed incident reports that explain what happened, the technical root cause, the impact, and the fixes. Its February 2026 BYOIP incident is a good example. - **Your actual experiments.** [Convex published the GTM channels it tested](https://stack.convex.dev/everything-we-tried-failed-except-for-this-one-weird-trick), including newsletter sponsorships, creator videos, Google and Reddit ads, and an AEO agency. It shared what happened, including the channels that produced poor results, and what it decided to stop doing. I would actively ask the marketing and product teams: “What do we know internally that our customers would find useful?” That could be your benchmark data, pricing logic, experiment results, product decisions, incident reports, or marketing numbers. Put that information in public. That is transparency I would actually use as a marketing tactic. ## **6. Connect product data to the systems your GTM team uses** Remember the trigger-based emails I mentioned earlier? Those only work if I know what the developer is doing after signup. A signup tells me almost nothing. I want to know whether that person created a project, made their first API call, connected an integration, invited teammates, used a paid feature, or is getting close to a usage limit. For example, if I run an API product, I could define a simple PQL as: Created a project + made 20 API calls + returned to the product three times in seven days. That account should not sit in the same CRM bucket as someone who signed up, looked around, and disappeared. I would send those events from the product into PostHog, Segment, or RudderStack, then push the useful signals into the CRM or marketing system. The workflow can then be very specific: - **First API call:** send the developer the next integration example. - **20 API calls in seven days:** move the account into a product-qualified segment. - **80% of the free quota used:** send the upgrade email. - **Three teammates invited:** flag the account for expansion. - **No API calls for 14 days after activation:** start a re-engagement sequence. - **Heavy usage from a company that matches the ICP:** create a sales task with the actual usage history attached. The signup is not the signal. What the developer does after signup is. Product-led sales teams use those actions to decide when to trigger onboarding, sales outreach, expansion, or re-engagement. If your CRM only says “John signed up,” you are missing the product signals that tell you whether John activated, expanded, hit a limit, or stopped using the product. ## **7. Boost thought leadership to the people already showing intent** This has worked particularly well for our enterprise clients. When a founder, CTO, or another executive already has a strong LinkedIn presence and consistently publishes useful content, I like putting paid distribution behind the posts that are already getting organic engagement. The audience should already know the company in some way. I usually retarget website visitors or upload a specific account or contact list, then use the executive’s post as a second touchpoint. For example, someone visits your website to read about your product, and a few days later they see the CTO explaining how their team approached a technical problem. **They are getting another useful piece of information from the company rather than another product pitch.** LinkedIn’s 2026 benchmark data from 161,256 ads across 211 companies found a **2.68% median CTR for Thought Leader Ads**, compared with 0.42% for single-image ads. The same dataset found that Thought Leader Ads work particularly well for mid-funnel education and trust-building. We have seen LinkedIn work particularly well for this type of retargeting with our clients. The post does the education, and the targeting gives it another chance to reach people who have already shown some interest. ## **8. Open-source the entry point to your product** Resend built **React Email** after running into a problem its own team had: building modern emails that worked across email clients. The library is open source, and developers can use it to build and preview their emails without paying Resend. Resend then provides the infrastructure to actually send those emails through its API. Browserbase has taken a similar approach with **Stagehand**. Stagehand is an open-source browser automation SDK that developers can run locally. When they need hosted browsers, session replay, captcha solving, observability, and production infrastructure, they can connect Stagehand to Browserbase. The open-source project should lead developers directly into the problem your paid product solves. React Email gets developers building email templates, which makes Resend a natural next step when they need to send them. Stagehand gets developers building browser automations, while Browserbase provides the hosted browsers and infrastructure to run those automations reliably in production. Give away the part developers want to use. Make the paid product the next step when they need to run it at scale. ## **9. Smaller event setups can work well** I have seen this work well with plenty of our clients. You do not always need a massive booth to get developers to stop and talk to you. Having engineers at the booth who can discuss the product, answer technical questions, and help someone build something gives developers a reason to engage. [**Solo.io**](http://Solo.io) used this approach at KubeCon with a hands-on lab where developers could try the software themselves. [Nearly 500 engineers completed the labs in one day, with each lab taking around five minutes.](https://instruqt.com/customer-stories/solo-io) That is how I think about developer events. I would rather put two strong engineers at a smaller setup and have them spend the day helping developers with technical questions and use cases. ## **10. Build into an ecosystem your ICP already uses** RapDev is a good example of this. It built integrations around Datadog, got them into the Datadog Marketplace, and used Datadog's existing customer base as a distribution channel. Within its first year, RapDev was getting [30 to 40 trial leads a month and $30K to $50K in monthly revenue](https://www.datadoghq.com/case-studies/rapdev/) from marketplace deals. I would look for the same opportunity in the tools my ICP already uses. If my customers spend their day in Datadog, GitHub, Linear, Slack, Terraform, or another developer platform, I would find a part of that workflow where my product can add value, build the integration, and work with the platform on distribution. The partner already has the developers. My job is to build something useful enough that they have a reason to put it in front of them. ### **Wrapping up:** Like I always say, marketing is 90% logic and 10% execution. These are some of the ideas and tactics that have worked for us at Hackmamba, or that I have picked up from some of the best technical marketers I have spoken to and learned from. The tools will change. The channels will change. The way developers discover products will change. The logic behind good marketing does not change nearly as fast. I would start there. Figure out what you are trying to achieve, why a particular tactic should get you there, and then use the tools and channels that help you execute it.I got hit by Google 2026 spam update.
— Hridoy Reh (@hridoyreh) August 24, 2026
All posts were 100% AI generated content.
See the fall of my AI content site: pic.twitter.com/K5IhaPn34A