Key takeaways
- Predictive prospecting ranks prospects by their likeness to past wins.
- It needs clean records of which deals were won and lost, and why.
- It improves focus, not message quality.
- Small teams can start with a written scoring rule before buying a model.
- Check the ranking against sellers' judgement and adjust it often.
How does predictive prospecting work
Predictive prospecting works by comparing new companies and people with the profile of your past wins and ranking them by how closely they match.
The inputs are fit data, such as industry, headcount, location and the technology a company uses, and timing data, such as hiring, new offices or new markets. A model, or a weighted rule, gives each prospect a score. Sellers work the list from the top, and the results feed back into the next ranking.
How does it improve lead generation
It improves lead generation by spending effort where a sale is more likely, so less time goes on accounts that were never going to buy.
It also makes targeting explainable. When a seller asks why an account is on their list, the answer is a set of traits shared with real customers, not a hunch. That makes it easier for marketing and sales to agree who to target.
What do you need before you start
You need a reliable record of past deals: which were won, which were lost and the traits of each account.
If you have only a handful of wins, write a scoring rule by hand from those deals. It is predictive prospecting in miniature, and much easier to correct.
- A CRM where won and lost deals are marked consistently.
- Company fields filled in for most accounts: industry, size and location.
- A written ideal customer profile to compare the ranking against.
- A sales team willing to work a ranked list and report back.
What are the limits of predictive prospecting
Its main limit is that it can only find more of what you already sell to, and it repeats any bias in your past deals.
A team that has only ever sold to software companies will get a ranking that favours software companies, even when a new market fits better. Company data also goes stale, so a ranking built on old records points at the wrong people. Treat the score as a starting order, and leave room for sellers to test new segments by hand.
How is it different from lead scoring
Predictive prospecting ranks companies and people you have not contacted yet; lead scoring ranks leads already in your system by fit and engagement.
The two work together. Predictive prospecting fills the top of the funnel with likely buyers, and lead scoring decides which of the people who respond deserve a seller's time first.
Where Pebbles Ai fits
- CapabilityProspecting & customer recordsAsk an assistant for companies that match your ideal customer, up to 50 per search, then find the right people at up to ten of them. The CRM, in Beta, keeps ideal customer profiles and a score badge on records.
- FoundationCompany IntelCompany Intel gives every assistant your business knowledge and brand voice, built from your website, company information and files, as the context for finding and writing to prospects.
Common questions
Is predictive prospecting the same as AI prospecting?
Not quite. Predictive prospecting ranks prospects by likeness to past wins; AI prospecting is broader and also covers research and drafting messages.
Can a small business use predictive prospecting?
Yes, in a simple form. List the traits your best customers share, score new prospects against them in a spreadsheet and work from the top.
How often should a predictive model be updated?
Whenever your deal history changes meaningfully, and at least each quarter. New wins and losses are the evidence that keeps the ranking honest.
About this page. Written by the Pebbles Ai team for people who run sales, marketing and growth. Last updated .
