A list filtered by company size and industry tells you who could buy. A list filtered by tech stack tells you who is ready to. The tools a company runs are the closest thing B2B has to a declared intention: they reveal fit (your product plugs into what they already own), pain (they run a competitor they may have outgrown), and timing (they just adopted an adjacent tool, or their current one is aging toward a renewal). Most of your market is not waiting to learn your category. They already run something, they are frequently unhappy with it, and a tech-stack list puts you in front of exactly those accounts: the ones already paying for a tool, already feeling its limits, and already a candidate to switch, integrate around, or fill a gap. This is how to build one.
Step 1: Decide which technologies signal a fit
Before you touch an enrichment tool, name the tools that predict a sale. This is the difference between a targeted list and a wide net with a logo filter on it.
There are three patterns worth building around, and a prospect can sit in more than one.
Flip each card to see the play a tech signal unlocks
Plays available: 3
A tech signal is only useful once you know which of three plays it unlocks: displacement, integration-fit, or gap-fill.
The competitor play converts best, because the buyer already understands the problem and has budget allocated to it. To find which tools predict your wins, interview ten current customers and ask what their stack looked like the quarter before they bought. The tools that show up repeatedly are your target signals.
Step 2: Build a company list from technographic data
A tech-stack list is only as good as the data underneath it, and most technographic data is shallow. Web crawlers detect what is visible in a website's source code: the chat widget, the analytics pixel, the CMS. That tells you a company "uses Salesforce." It does not tell you which edition, how much they spend, when they implemented it, or whether they are about to leave.
In Clay, build the list with the Companies by product usage with HG Insights source, which finds companies by the tools they run, including back-of-house tools no website crawler sees. HG Insights reads the company instead of the website: it processes business documents, SEC filings, contracts, RFPs, earnings calls, and job postings to recover spend, edition, install date, and usage depth that scraping never sees. Add a column, add enrichment, select "Companies by product usage with HG Insights," and start with 10 to 50 companies (or Sandbox mode) so you validate targeting before spending credits.
You can define the technology three ways, and which one you pick changes the list.
Switch the search mode and watch how wide the net gets
salesforce.comNarrowReturns: Companies using any product from that vendor
Sample results
Use when you want everyone in a specific vendor's ecosystem.
Searching by product category casts the widest net for the same per-company credit cost as a single product, so start broad when function matters more than vendor.
If a tool you want is not in the product-name dropdown, narrow first by vendor domain and category; if it still does not appear, confirm it exists in HG's database at discovery.hgdata.com before assuming the list is empty. Set a recurring schedule so new adopters flow in automatically rather than going stale the day you build it.
Step 3: Layer firmographic filters to turn it into a prospect list
A tech-stack search alone returns a technology's entire install base, which is a market, not a list. Firmographics cut it down to the slice you can actually sell to.
Set an employee-count range that matches your ICP (say 100 to 1,000 for mid-market), a revenue floor so every account has budget, the industries where you have product-market fit, and geography down to country, state, or city. One filter matters more than it looks: toggle headquarters versus subsidiary locations. Targeting decision-making locations rather than any office that happens to run the tool is what keeps a regional sales rep from chasing a branch with no buying authority.
Firmographics tell you who to target; technographics tell you how to approach them. The accounts that sit in the intersection, right size and right tech, are the only ones worth a rep's time.
Step 4: Filter for the specific play, not just the technology
Owning a competitor's product is a signal. Owning it badly is an opening. The accounts most likely to switch are not the ones that simply use a competitor; they are the ones whose usage is thin, aging, or about to renew.
HG Insights gives you the product first-verified date and the total product signals count. Renewal timing is not a field you pull; you infer it from install age and a falling signal count.
Stack the conditions and watch the install base collapse to the switch-ready
Switch-ready: 0 of 8
Stacking install age, low usage, and renewal timing turns a competitor's entire customer base into a short list of accounts actually in a buying window.
For an integration-fit play, run the opposite logic: target recent adopters of a complementary tool. A company that just brought on Gong or Chorus is investing in sales technology and is a candidate for the next layer. For a migration play, hunt outdated versions of a platform approaching end-of-life, the accounts already in evaluation mode whether they have called a vendor or not.
This is the depth that changes a GTM motion rather than just cleaning the data feeding it.
“We consolidated three vendors into Clay and started enriching data points that didn't exist in any traditional database. Our reps went from starting every conversation cold to knowing exactly who to call and what to say.”
Step 5: Enrich contacts at the accounts that survived
A company list is not a prospect list until it has people on it. Once your accounts are filtered, run Clay's contact enrichment to find the right roles at each one, then run a waterfall to recover work emails and direct phone numbers from professional profiles and public web data across more than 150 sources, checking the next provider whenever one comes up empty.
Match the role to the play. The same tech signal lands differently depending on who reads it.
Tap a stakeholder to see how the same stack reads to them
Salesforce + HubSpot + Outreach
One detected stack, three different pitches
What IT sees
Managing middleware between three systems; security surface area
Opener angle
Our unified platform removes the integration layer you're maintaining between Salesforce, HubSpot, and Outreach.
The same tech stack is three different pitches depending on whether IT, marketing, or finance reads it, so multi-thread the signal, don't repeat it.
Step 6: Personalize off the tech signal
The reason you built the list this way only pays off in the message. A name and a title produce "I noticed you work in marketing." A tech stack produces a sentence the prospect could not have received from anyone who did not do the research.
Specificity and timing are what move a 2% reply rate toward 20%. Reference the exact tools, name the integration challenge those tools create together, and point at the gap your product fills. Feed the enriched stack fields into a Clay AI column with a prompt that turns data into a first line.
You are an SDR writing the first line of a cold email.Inputs: company name, detected tech stack (list), our product category, the play (displacement / integration-fit / gap-fill).Write one sentence (max 25 words) that names a specific tool or tool combination from their stack and the concrete friction it implies. No greeting, no pitch, no exclamation marks. Reference the stack, not the company size or industry.If the play is displacement, reference likely frustration with the named competitor near renewal. If integration-fit, name how our category connects two tools they run. If gap-fill, name the tool that's conspicuously missing.
A line like "I saw your team uses Salesforce Enterprise plus Marketing Cloud but still runs Mailchimp for some campaigns" works because it proves you looked, and it states a problem the reader already feels. That is the whole point of building the list by stack instead of by size.
Common failure modes
Most tech-stack lists fail in predictable ways, and all of them are avoidable before you hit send.
Expand each mistake to see the fix
Every tech-stack list failure traces back to one of four fixable habits: stale data, no firmographic cut, treating install as intent, or generic outreach.
The throughline is that a tech-stack list is a precision instrument. Build it with depth, cut it with firmographics, time it with install data, and the message writes itself off the stack you already know.