Increasing inbound leads from LLM recommendations
The company had a strong product with a founder doing all of the selling. We ran their paid search and rebuilt their sales calls, and both started producing more. The bigger surprise came from the website. Once we made it easy for AI assistants to read and trust, the visitors they sent became the best leads the company had, and we built the next marketing plan around that.
B2B software company // Client story
50%
of ad spend redirected
About half the budget had been reaching people who would never buy. We moved it to real buyers.
1 in 6
deals from AI referrals
Counting new deals with a recorded source. AI visitors also converted better than any other inbound channel.
5 weeks
to sign the biggest deal
We rebuilt the pitch and trial plan for their largest open deal. It signed five weeks later.
Where the work started.
The ad reports showed more than half of visitors engaging. When we traced what those visitors did next, about 3 in 100 got in touch, and half the budget was reaching people who would never buy. On sales calls the demo came early, before the buyer had said much about their own situation. The website gave AI assistants very little they could read or repeat.
//paid_search
Fixing the ads, starting with the numbers
We ran their Google Ads end to end and gave the account a full overhaul, starting with the measurement. Conversion tracking had broken in a site rebuild, the ads were showing on the display network where few buyers were looking, and half the budget was going to people who'd never buy. We rebuilt the tracking, moved the spend to search and tightened the keywords. The outcome was a large increase in inbound for the targeted product.
//discovery_and_demos
Asking before showing
We sat in on their sales calls and went through more than 25 recordings. The same product was being sold to two very different buyers, each with its own reason to buy. We rebuilt the discovery call so the team understood the buyer before opening the demo, and used what we heard to sharpen the ideal customer profile quickly. From there we set a technical demo sequence the team could repeat, and deals started moving faster.
//ai_discoverability
A website AI assistants can read
The goal was to increase AI referrals. We changed both the site's code and its copy. In the code, we added the structured data search engines and AI assistants look for, and fixed a setting that was hiding a key page from search. On the pages, we rewrote the headings and answered the questions in prose we'd seen buyers use. We also built a tracker for how often AI assistants named the company. The outcome was a significant (and continued) rise in visits from AI referrals. These visitors continue to be far more likely to request and attend demos.
//partnerships_and_content
Getting recommended more often
Once we improved the site for AI referrals, the question was how to get recommended more. AI assistants recommend companies that other people talk about, so we planned a partnerships program to get the founder speaking to audiences that matched their buyers, with online announcements. We paired it with a plan to create and share useful material far more often. Website traffic rose +10% in the first quarter and continues to expand.
What the team was left with.
They have a clear ad strategy, a discovery-first sales process and a website AI assistants are recommending. AI referrals are now their best leads, with recommendations continuing to climb, and the website-to-call ratio climbing.