How Gravte Generated 43 MQLs in 90 Days for a Healthcare AI Startup with LinkedIn Ads

Client

101genAI — a San Francisco-based healthcare AI startup helping healthcare companies adopt AI in a safe, compliant way. Their platform audits clinical and operational interactions to maximize care quality, patient satisfaction, and financial outcomes.

Industry

Healthcare AI / HealthTech B2B SaaS

Outcome

43 leads in month 3, 8% lead-to-SQL conversion rate, 5.4x month-over-month lead growth across the 90-day campaign

Scope

ICP research, persona development, LinkedIn paid ads — strategy, creative, targeting, A/B testing, and ongoing optimisation

101genAI came to us at MVP stage. The product was built. The use case was clear — helping healthcare organisations adopt AI without cutting corners on compliance or care quality. What they didn’t have yet was a repeatable way to get in front of the right people and turn that interest into pipeline.

Their target was a specific kind of buyer: healthtech startup founders and operators at companies actively exploring or implementing AI, based in the major startup hubs across the US — San Francisco, New York, Boston, Austin, Seattle. Not a broad audience. A very specific one. Which meant the strategy had to be precise from day one.

The challenge wasn’t just lead generation. It was building a paid acquisition engine from scratch — no prior campaign data, no creative history, no conversion benchmarks — and making it more efficient with every passing month.

Problem

  • 101genAI was at MVP stage with zero existing paid acquisition infrastructure — no campaigns, no conversion tracking, no historical data to build from
  • The ICP was narrow and high-intent: healthtech founders and decision-makers who would only respond to messaging that spoke directly to their compliance and quality concerns — generic ad copy wouldn’t work
  • Healthcare AI is a trust-sensitive category — buyers are cautious, the sales cycle is longer, and vague or overpromising creative gets ignored fast
  • LinkedIn targeting for a niche audience like this requires careful layering of job titles, company size, industry verticals, and geography — blunt targeting burns budget quickly
  • As an early-stage company, 101genAI needed leads that were genuinely sales-ready — volume without quality would waste the team’s limited bandwidth

Solution

The strategy was built in three iterative phases. Each month we had more data than the last, and we used every bit of it.

Month 1: ICP research, persona development, and campaign foundation

  • Ran deep ICP discovery sessions with the 101genAI team to map out who their best-fit buyers actually were — their roles, the problems keeping them up at night, how they talked about compliance and AI adoption internally
  • Built a detailed ideal customer persona: healthtech startup founders and VP-level operators at companies in active AI adoption or evaluation, based in SF, NYC, Boston, Austin, and Seattle
  • Set up LinkedIn Campaign Manager from scratch — campaign structure, budget pacing, bid strategy, and conversion tracking wired to the right downstream events
  • Launched an initial set of ad creatives designed for A/B testing — varying hooks, visual formats, and CTAs to establish what resonated with this specific audience
  • Kept targeting broad enough in month one to gather signal before tightening, rather than over-constraining the audience before we had data

Month 1 result: 8 leads. Baseline established. Early signal on which creatives were earning clicks and which weren’t.

Month 2: Campaign optimisation from real data

  • Analysed month one performance at the ad level — paused underperforming creatives, doubled down on what was working
  • Refined audience targeting based on engagement signals — tightened job function and seniority filters, cut geography segments that weren’t converting
  • Iterated on ad copy to sharpen the value proposition — moving away from broad AI messaging toward specific compliance and care quality pain points that the ICP actually cared about
  • Improved landing page and lead form alignment with ad messaging to reduce drop-off between click and conversion

Month 2 result: 13 leads — a 62% increase over month one. Lead quality noticeably improved, with more SQLs in the mix.

Month 3: Scaled targeting and creative maturity

  • With two months of conversion data, we had enough signal to significantly tighten targeting without risking reach — layering company headcount, funding stage, and specific healthtech sub-verticals
  • Developed a new round of creatives informed by what month one and two had taught us about the audience — stronger hooks, clearer differentiation, formats that matched how this audience actually consumes LinkedIn content
  • Introduced retargeting for engaged prospects who had clicked but not converted, keeping 101genAI top of mind through the consideration phase
  • Optimised bid strategy as the campaign algorithm accumulated enough data to exit the learning phase and spend more efficiently

Month 3 result: 43 leads — a 231% increase over month two and 5.4x month one. Lead-to-SQL conversion rate hit 8%.

Outcomes

  • 43 leads in month 3, up from 8 in month one — a 5.4x increase over the 90-day period
  • 8% lead-to-SQL conversion rate — well above typical B2B LinkedIn benchmarks, reflecting the quality of targeting and creative-to-audience fit
  • Leads were genuinely sales-qualified — healthtech founders and decision-makers actively evaluating AI adoption solutions, not just curious clicks
  • Built a repeatable paid acquisition engine with documented targeting frameworks, creative playbooks, and conversion benchmarks that compound in value over time
  • Gave 101genAI’s sales team a predictable, high-quality inbound pipeline for the first time since launch

The growth curve here wasn’t an accident. It was the result of treating the first month as a learning investment, not a performance expectation. When you’re launching paid acquisition for a niche audience with no prior data, the most valuable thing you can do is set up the infrastructure correctly, gather real signal, and iterate fast.

By month three, we weren’t just running ads — we were running a system. And a system built on clean data, sharp creative, and a well-understood buyer compounds. That’s what 43 MQLs in 90 days looks like.

Become the obvious choice

Let's talk