Welcome folks! 👋

This edition of The Product-Led Geek will take 5 minutes to read and you’ll learn:

• How AI assistants have become new gatekeepers in developer tool selection, fundamentally changing discovery patterns.

• The four key elements that make your SDK discoverable to AI.

• Why radical clarity matters in becoming the default recommendation in AI-dev workflows

Let’s go!

TOGETHER WITH

Delve helps fast-growing companies get compliant with an AI-automated platform. SOC 2, HIPAA, and any other certification in just 15 hours.

The proof is in the pudding:

• Lovable → From zero to fully SOC 2 compliant in just days

• Cluely → Unlocked millions in enterprise contracts within just days

• Bland → Switched over, got SOC 2 compliant and unlocked $500k ARR within 7 days

If you’re suffering elsewhere, they’ll even migrate you off another platform all included

Product-Led Geeks readers get $1,000 credit using code PRODUCT1KOFF + Free Airpods Pros when you get compliant

Getting compliant in days >>> quarters.

Please support our sponsors!

GEEK LINKS

GEEK OUT

How AI Assistants Choose Your SDK (and Why It Matters)

The first Google search result doesn't matter anymore.

When developers need to add email, payments, or search to their app, they're asking AI assistants - and those assistants are finding options, and they're making choices.

They pick one solution and write integration code on the spot.

This matters because AI assistants don't just search - they recommend.

Every public trace of your product - docs, code samples, package names, Stack Overflow answers - shapes whether you show up in these critical moments.

Get it right, and AI becomes a powerful distribution channel.

Get it wrong, and you're invisible.

So how do you actually get chosen by these new gatekeepers?

How do you become the answer in a copy-paste AI world - and not just more noise in the long tail?

The New Dev Funnel

The old developer funnel was forgiving:

• Show up in search results

• Get compared against alternatives

• Maybe make the shortlist

• Convert some percentage who try you

The new AI funnel is much less so:

• You either show up in the model’s answer - or you don’t.

• The code snippet runs right away - or it fails.

• If those go smoothly, you’re in the project (and much more likely to get adopted at scale).

Typically developers evaluate multiple tools before making a choice - reviewing documentation, comparing features, and reading community feedback.

While AI assistants are changing how some integrations happen, their impact varies by context. An experienced developer isn't going to switch from Stripe to an alternative just because an LLM suggested it.

Years of experience with battle-tested solutions, existing team knowledge, and proven reliability matter far more than AI recommendations for critical infrastructure choices.

Though as I’ve written about in past issues, in vibe mode, even those decisions are delegated.

In typical professional-dev workflows, where AI assistants are having more impact is on specialised but non-critical components - like rich text editors, email composers, analytics integrations, or media handling libraries.

These are substantial features where developers might be more open to trying the assistant's suggestions, especially if the solution seems well-documented and maintained.

The key is understanding where in the stack these new discovery patterns matter most.

How to Make Models Choose You

AI models are pattern matchers, not strategic thinkers.

They pick up on repetition, clarity, and correlation with successful outcomes.

Your job is to be consistently, boringly clear about what you do and how to use you.

• Names and identities: Assistants attach meaning and domain from names, one-liners, and context. If your name is generic or inconsistent, you lose.

• Docs structure: Page titles, headers, and how you group tasks matter much more than clever copy or fancy landing pages.

• Examples: Models care about the shape and placement of your code snippets, far more than any marketing speak on the page.

• Consistency: Repeating the right phrasing, using the expected nouns and verbs, and organising for action (not for marketing) all add up.

When your material is scattered, ambiguous, or buried, assistants guess - and those guesses are where you lose out to a competitor whose presentation just happens to make more sense to the model.

Making Your Tool Discoverable

The first rule of AI discovery is brutal simplicity. Here's what actually works:

1. State Exactly What You Do

Use a phrase that signals exactly what you do.

Add use it everywhere.

Every surface needs the same clear one-liner.

For example:

Put this everywhere - homepage, README, package registry, documentation.

No creativity, no variation.

Just the plain truth, repeated.

2. Write Docs for Real Tasks

Models anchor on headers and page titles.

If you want to be the answer for "send transactional emails," you need exactly that phrase as a page title.

Your docs should read like a task menu:

• "Process Stripe webhooks" (not "Webhook Integration")

• "Upload files to S3" (not "Storage Options")

• "Send bulk emails" (not "Email Features")

Include:

1. One sentence saying when to use it

2. Working code snippets that solves the exact task

3. Zero fluff

While most dev tools keep their docs focused on technical content, any documentation that delays getting to working code - whether through verbose explanations, unnecessary context, or marketing copy - risks losing both human developers and AI assistants to more direct alternatives.

The best performing docs get straight to the solution, with just enough context to use it correctly.

Think GitHub README style: problem, solution, code.

3. Own Your Comparisons

If you don't write your own comparisons, someone else will - and that's what models will quote.

Take control:

1. Create a "/compare" page that ranks high in search:

• yourproduct.com/compare/competitor

• yourproduct.com/alternatives

2. Write direct feature comparisons:

"Our search handles typos and synonyms out of the box, while Elasticsearch excels at exact text matching and log analytics"

3. Build comparison tables models can parse:

4. Add contextual signposts in docs:

"Need [different use case]? Consider [alternative approach]. Need [your core use case]? You're in the right place."

If you skip a “vs Competitor” or “Alternatives” section in your docs and READMEs, someone else will write it for you somewhere…and LLMs may end up quoting or paraphrasing outdated, incomplete, or even inaccurate information.

Why do this?

Structured, honest comparisons help models give realistic, trustworthy answers - and help developers make the right decision fast.

Assistants and LLMs are much more likely to pick up the facts and framing you provide than to guess or paraphrase rumours from the open web.

Own the narrative about where you fit in the market.

If you make competitor comparisons easy to find, specific, and credible in your own docs and guides, both humans and AI assistants will be far more likely to understand - and represent - your strengths accurately.

The New Rules of Discovery

Getting found by AI assistants is very much analogous with getting found in the search engine era.

Prompts are the new keywords.

In the past you would strive to authoritatively own keywords.

Today you do the same for prompts.

The way to do that is through radical clarity and consistent signals:

1. Name things clearly and consistently

2. Write docs that match real developer queries

3. Own your comparisons before others do

4. Make your examples unmissable

5. Keep your identity unified across all surfaces

The tools have changed, but the fundamentals haven't:

Be clear, be honest, be helpful. AI just amplifies the returns on getting these basics right.

Next week, I'll dive into implementation: how to improve the chances that the code AI assistants generate with your SDK actually works. Everything from perfect example design to handling breaking changes without losing developer trust.

But for now, remember: You can't trick an AI assistant into recommending you. But you can make it obvious that you're the right answer.

THAT’S A WRAP

Before you go, here are 3 ways I can help:

Take the FREE Learning Velocity Index assessment - Discover how your team's ability to learn and leverage learnings stacks up in the product-led world. Takes 2 minutes and you get free advice.

Book a free 1:1 consultation call with me - I keep a handful of slots open each week for founders and product growth leaders to explore working together and get some free advice along the way. Book a call.

Sponsor this newsletter - Reach over 7600 founders, leaders and operators working in product and growth at some of the world’s best tech companies including Paypal, Adobe, Canva, Miro, Amplitude, Google, Meta, Tailscale, Twilio and Salesforce.

That’s all for today,

If there are any product, growth or leadership topics that you’d like me to write about, just hit reply to this email or leave a comment and let me know!

And if you enjoyed this post, consider upgrading to a VIG Membership to get the full Product-Led Geek experience and access to every post in the archive including all guides.

Until next time!

— Ben

PS: Thanks again to our sponsor: Delve