Today’s plays › All plays › Prospecting & research
Have AI fix the three inputs your account score depends on, then let the score pick your accounts
Picked and checked by Olga Pearson, SalesUpLevel
Get tomorrow’s play in your inbox. Free ↓
Reported in Clay's case study: top-tier enterprise accounts now close at a 22% higher rate and carry a 50% larger deal size than accounts that were scored wrong. No time frame is given.
I keep seeing teams pour hours into accounts that look A-tier on paper, then find out the score was built on bad industry and location data. That is the quiet leak.
Owen Chandler at AlertMedia (in a Clay case study) faced it head on. Their account score hangs on a few inputs: how many locations a company has, where they are, and its industry. Data vendors kept getting those wrong. Their GTM engineer had an AI agent re-check each company's industry, pulled location counts from several data providers, filled missing legal names and LinkedIn pages, and wrote it all back to Salesforce on a schedule, so the score tells the team where to spend time.
By Owen Chandler (AlertMedia), in a Clay case study, GTM Engineer, AlertMedia · Case study (Clay customer story) ↗ · Shared by Clay on LinkedIn Apr 30, 2026 (page has no date)
Run it before your 10am
- List the 3 to 5 fields your account score depends on most.
- Spot-check 50 accounts: how often is each field wrong or empty?
- Have an AI agent re-check the worst field from the company's own website (for example, industry).
- Fill counts like locations or headcount from more than one data source, and keep the best match.
- Write the fixed fields back to the CRM on a schedule, not once.
- Compare win rate and deal size for top-tier accounts scored on clean data versus the old score.
Here is a company: [name, website]. From their own website and public sources, tell me: 1. Their industry, using only this list: [your industry list]. Give the source. 2. How many physical locations they run, and in which countries. Say if this is an estimate. 3. Their legal company name and LinkedIn company page. If a source disagrees with what we have in the CRM ([current values]), show both and say which you trust and why.
If you lead a team
- Pick the one input sellers complain about most and fix that first.
- Report win rate and deal size by score tier every quarter.
Watch out for
- Vendor case study; Clay is selling enrichment.
- The 22% and 50% compare against 'improperly scored' accounts, not a before-and-after period.
- Only worth it if your score is built on a few inputs you can name.
- What I'd have my team try first: pick the one score input sellers complain about most, fix that field on 50 accounts, and only then talk about re-ranking the whole book.
Source: Clay customer story 'AlertMedia' (shared by Clay on LinkedIn, Apr 30, 2026). Case study: customer AlertMedia, vendor Clay. Results are as reported in the vendor case study; we did not measure them.
Get tomorrow’s play before your first call
Free. One sourced play a day.