JournalSignalsNo. 13
Signal Stacking for B2B Sales: 5 Plays You Can Build
Combine independent buying signals into a clear account decision. Five practical plays, a worked example and a free signal-stacking CSV worksheet.
Matija Nikolić9 min read9 sections

Contents
Published by Perpetual
Signal stacking means combining distinct observations about the same company to decide whether it deserves attention and what to investigate next. A useful stack connects account fit, a relevant change and evidence you can verify. It does not turn several uncertain clues into proof that someone has budget or wants to buy.
For a small B2B sales team, start with a question your offer can help answer. Then collect the few observations that change your decision. A long feed of activity is less useful than a short record explaining why this account, why this person and what remains unknown. This guide gives you five plays, a worked example and a free signal-stacking worksheet.
Separate fit, events and direct interest
These three kinds of information do different work. Fit tells you whether your offer could serve the company: the market, operating model, geography and relevant responsibilities. An event tells you something changed, such as a published role or an announced regional launch. Direct interest tells you what someone actually asked or did in a relationship you can identify.
| Layer | Example | Decision it supports |
|---|---|---|
| Fit | A B2B software company serves the region your service covers. | Whether the company belongs in your audience. |
| Relevant event | A current job posting describes building a regional sales team. | Which work to research and whether the timing deserves a look. |
| Direct interest | An identified contact asks how your research service handles that region. | How to answer the request and qualify a conversation. |
A company can fit perfectly without a relevant current event. It can also produce many events while remaining a poor fit. Check eligibility before investing in the stack, and keep direct requests separate from inferred interest. Our buying-signal library covers the underlying signal types.
There is no universal number of signals that makes an account ready for outreach. The test is whether the combined evidence changes a defensible decision. Unify's practitioner guide likewise distinguishes collecting signals from deciding which ones should activate a workflow. The plays below are our research framework, not measured conversion rankings.
Check independence before adding evidence
Three job sites carrying the same vacancy describe one opening. A company announcement and two articles repeating it describe one announcement. Keep useful links, but group them under the underlying event rather than awarding extra weight for each copy.
Independence concerns the observation, not just the domain. Two different pages on an employer's site may describe separate decisions: an expansion announcement and a local hiring brief, for example. Two different publishers may be repeating the same press release. Ask what new fact the second source contributes.
A job description containing the words “new territory” adds detail to the vacancy; it is not independent confirmation that a territory launched. A regional launch announcement can provide that separate confirmation. This distinction keeps a useful description from becoming an inflated stack.
Record the event date when known, the date you first found the source and the last time you checked it. A page discovered today may describe an older event. If timing is unknown, record it as unknown instead of making the account look fresh.
Five signal stacks to investigate
Each play begins with an eligible account. The additional research can be manual or supplied by another source; the table does not imply that one product automatically collects every signal.
| Play | Evidence to combine | What to investigate | What remains unknown |
|---|---|---|---|
| 1. Hiring with a specific mandate | A current role describes a new responsibility; a separate company page explains the initiative it supports. | Who owns that work today and how your offer supports it. | Whether the team wants outside help. |
| 2. RevOps and CRM context | A RevOps posting names routing work; a separate current source establishes the relevant CRM context. | Whether your routing offer fits the documented environment. | Whether a migration, replacement or purchase is planned. |
| 3. Regional hiring and a market launch | A regional sales opening and a dated company announcement about entering that market. | How local prospect research will be organized. | Launch scope, ownership and supplier budget. |
| 4. A leader and distinct team openings | A verified leadership appointment and separate active roles in the function. | Which responsibilities the leader actually owns and whether the roles support the same initiative. | The leader's priorities and willingness to consider vendors. |
| 5. Public change and known engagement | A relevant public event and a separately recorded interaction from an identified contact. | How to respond to that person's actual interest in the context of the event. | Whether the engagement represents a funded evaluation. |
For play one, do not count the role and its own description twice. The separate initiative source is what adds evidence. For play two, a tool mentioned only as desirable experience does not establish that the company uses it. If the CRM context cannot be verified, keep the play as a research task.
For play five, an anonymous visit matched to a company does not identify the visitor. A named contact downloading a resource also does not authorize you to claim that the whole buying group is evaluating. Respond within the relationship and permissions you actually have. The distinction between observable hiring and modeled research activity is covered in hiring signals vs. intent data.
Turn a stack into one account decision
Keep one record per company and connect its underlying events to that record. More sources should improve the explanation, not multiply recipients or create duplicate campaign entries.
- Check fit. Confirm the offer, segment and geography match. Check existing customers, active opportunities, ownership and outreach exclusions in the systems your team uses.
- Describe the observations. Use narrow factual sentences with sources. Keep your interpretation in a separate field.
- Check freshness and independence. Reopen the underlying pages. Group copies, and identify what each distinct observation adds.
- Identify a responsible person. Research who owns the work, rather than choosing whichever executive has an available email.
- Choose an action. Contact, investigate, wait or pass. Add the reason and, when waiting, the event that would justify another look.
A numerical score is optional. If you use one, describe its inputs and treat it as a prioritization aid until tested. A score of 80 is not an 80% purchase probability simply because the interface displays a percent sign.
Waiting should have a condition. “Recheck when the regional owner is identified” is actionable; “nurture forever” is not. Passing is also useful when the company does not fit or the connection to your offer cannot be explained.
A worked account and a counterexample
The companies and details in this example are fictional. Assume you sell account-research services for B2B teams entering the UK.
Harbor Example announces a UK launch on its own newsroom. Its current careers page separately lists a regional account executive role with responsibility for developing UK accounts. Those are two observations about the same initiative. A current team page identifies the regional sales leader.
The supported conclusion is that a UK sales initiative is worth investigating. You still do not know whether research is central, local or outsourced. Your question can make that gap explicit:
Your UK launch announcement and regional AE opening caught my eye. We help B2B teams prepare account research for new territories. Will the UK team build its own account research, or will that stay with your central sales team?
Meadow Example has three search results for a remote sales vacancy. All three resolve to one old opening that no longer appears on the employer's board. The company also falls outside the markets your service covers. More links do not rescue the fit or freshness. Pass instead of drafting a similarly confident email.
These accounts show why a threshold such as “three signals means send” fails. The first account has fewer records and a better-supported question. The second has repeated records and no current reason to approach.
Test the play without hiding the work
Choose one offer and one play for an initial batch small enough to inspect in full. Keep companies that fail qualification in your record. Otherwise the stack can appear efficient simply because rejected accounts disappear from the denominator.
Track reviewed companies, eligible companies, contacted companies, positive replies, qualified meetings held and total cost, including research time. Record why accounts were held or passed. The campaign scorecard explains these measures and their denominators.
If you compare a stacked play with a single-signal play, keep the offer, segment and follow-up policy comparable. Record any selection differences before interpreting results. A small successful batch is a reason to keep learning; it does not establish that adding signals caused the outcome.
Download the signal-stacking worksheet
Download the free signal-stacking worksheet (CSV). It includes all five plays, a duplicate-event example and a blank account record. Replace the sample observations with evidence you verify. It is a planning worksheet, not a product import or a validated scoring model.
Where Perpetual fits
Perpetual starts from public hiring activity, qualifies companies against your offer, selects one relevant person with an explanation and drafts an opener grounded in the evidence. Its current process does not imply automatic collection of every external or first-party source described in this guide. Each opportunity carries the limits of what its evidence proves.
Email runs from your connected mailboxes within the sending policy you configure; replies and opt-outs stop follow-ups. LinkedIn steps remain manual tasks. Start with the quickstart guide to see how the hiring-based process works.
Frequently asked questions
How many signals should I stack?
Enough to improve a specific decision. Two relevant, distinct observations can be more useful than ten repeated or unrelated events. There is no universal threshold that establishes purchase readiness.
Can two sources confirm the same event?
Yes. They can improve confidence that the event happened, but they should remain one event in your account record. Distinguish corroborating sources from separate business developments.
Should a signal stack trigger an email automatically?
Only after the account, person, evidence and sending rules meet your campaign policy. A weak or incomplete stack can trigger research or a hold instead.













































