Case study · Positioning and funnel diagnosis
The client asked for 5 channels. I gave him one, and a list of things to fix first.
01
The situation
The founder reached out in December. His co-founder was stepping back to join an accelerator for another project, and that co-founder had been the one generating most of the growth. The company was compounding at about 10% month over month and he wanted to protect that number without the person who was producing it.
He came with a list. Technical blogs, forum activity, cold email, social media, paid campaigns. He also wanted to start selling to enterprise, because in his words there's huge money there even with one client.
The budget was 5 hours a week.
02
Why I turned most of it down
5 channels at 5 hours a week isn't a marketing program. It's a rounding error spread across 5 surfaces, and none of them get enough attention to produce a signal you can actually read.
So I proposed 2 channels, then argued myself down to one. The pitch was simple: pick the smallest thing we can do properly, run it for 30 days, and use the result to decide whether to scale up or stop. No minimums, no lock-in. Easier to start narrow than to stay stuck in the planning phase.
He agreed. We started with outbound.
Then I looked at the website.
03
What the audit found
I ran a full audit, not an SEO audit. Everything a visitor would actually hit.
The navigation was broken in a way that's hard to describe without sounding like exaggeration. 4 of the 8 items under the main solutions menu pointed at pages that had nothing to do with their labels, because those pages had never been built.
The footer introduced a second set of pages that didn't appear in the top navigation, some of them duplicates of pages that did. The primary "Buy Now" call to action dropped users onto a separate dashboard subdomain with a different layout and no way back. The logo didn't link home.
The top-level menu label was "Pricing," while both direct competitors used "Products" or "Proxies" and organized by IP type. This company organized by use case, under a pricing label, with half the destinations wrong.
I sent the founder the list and one sentence that reframed the engagement:
At the current state of the site, we'd be pouring people into a broken funnel.
He agreed to pause outbound and let me fix the foundation first.
04
The constraint nobody had named
The second problem was harder, and it wasn't about execution.
The company's marketing was written as though it could serve every use case the large competitors served. It couldn't. The infrastructure is US-only static residential IPs. No international coverage, no rotating residential pool at scale, no managed scraping products or prepared datasets. Against the category leaders, that's a losing comparison on almost every axis they were implicitly inviting.
But the same constraint is an advantage in specific places, and nobody had written it down:
- The IPs are dedicated, so they don't get burned by another customer's overuse
- The IPs are static, so a session doesn't rotate mid-login and kick a user out
- Unlimited bandwidth at a flat rate, which is genuinely competitive when the workload is high volume and US-focused
That's not a proxy company competing with the giants. That's a proxy company that's very good at a narrow set of jobs.
So instead of writing an aspirational positioning, I proposed we find out where they were already winning. There were existing customers using the product for things nobody at the company fully understood. I built a customer interview questionnaire designed to surface 3 things: why they chose this provider, what the product does well that the company doesn't know about, and what's missing that could be closed without major engineering investment.
Find the segments where the constraint reads as a feature. Market those. Stop pretending about the rest.
05
What I built
- Information architecture. Restructured the menu around how buyers actually search. "Pricing" became "Proxies," organized by IP type with specialized tools grouped separately. "Solutions" became "Use Cases." Fixed the broken destinations and a hover bug that was closing the menu mid-navigation.
- Use case page template. Built the SEO and SERP tracking page end to end, copy and layout, as the pattern for the rest. The rule I set: page copy leads with the use case problem, not with generic product advantages.
- Trial flow. The existing trial process required emailing support with a specific subject line and waiting 24 hours for approval. Built a proper signup form, linked from every use case page.
- Outbound infrastructure. Roughly 5,000 prospects in Apollo, segmented by inferred use case rather than by industry alone. Targeting was technical decision makers at companies under 200 employees.
- 4 email sequences, one per use case. Emails get shorter as the sequence progresses. No repeated claims across steps. Every sequence closes with a no-oriented question, which gives the recipient the easy exit and produces replies from people who would otherwise ignore a 4th email. Each sequence leads with the differentiator that matters for that specific job.
06
What I'd tell another founder in this position
The instinct when growth is at risk is to add channels. More surfaces, more chances something works.
Usually the opposite is true. If the funnel leaks and the positioning promises things the product can't deliver, every additional channel just increases the rate at which you disappoint people. You spend more, you learn less, and the results are unreadable because nothing was isolated.
Fix the funnel. Name the constraint honestly. Find the customers who are already happy and figure out why, before you go looking for new ones.
Then add channels.