The Competitive Displacement Signal Reference

This is a lookup document for anyone running competitive displacement: founders selling their own product, account executives, SDR leads, and partnerships teams who want to reach accounts that already run a rival and are ready to leave it. Most public pages on this topic list four or five signals then pitch a monitoring tool, and none separate "uses a competitor" from "ready to leave a competitor." This reference defines each displacement signal, states what it proves and how it produces false positives, gives the productive timing window, and shows how to source and attribute it from open evidence. Jump to one row, act, and come back.

Why "uses a competitor" is not a lead

An account running a competitor is a static footprint, not switching intent. The lead is the dated event on top of the footprint: a renewal window, a champion move, a product outage, a dated complaint, or growth the incumbent cannot support.

The reason this matters is that raw footprint lists barely differentiate reps. In Refolk's index of professional profiles, the number of US professionals listing Salesforce as a skill and the number listing HubSpot are almost even, which means a "who uses a competitor CRM" list is close to symmetric across the category. Everyone can build that list. The advantage lives in the switchable subset.

These are skill-footprint counts from Refolk's index, a proxy for how many people work with each tool, not a count of company installs. The point is the ratio: 1.07 to 1. A displacement list built on footprint alone splits the market roughly in half and tells you almost nothing about who will pick up the phone.

The displacement signal library

Each row below is a signal you can look up on its own. "What it proves" is the switching-intent claim you may make from it. "How it lies" is the false positive that produces wasted touches. "Window" is when the signal is productive.

Footprint signals: proof of use, not intent

Tech-stack detection. Detected from script tags, DNS records, IP ranges, and cookies. Script tags surface front-end and marketing tools; DNS and IP surface infrastructure. What it proves: the tool is or was present. How it lies: a script tag left behind after a tool was removed reads as a live install. Window: none on its own; it is a base-list signal, not a timing signal. Source: tech-detection services cross-referenced with a source count.

Job postings listing a competitor as a required skill. When a company lists a tool as a required skill in an open role, that signals active internal adoption even when the tool never touches a public page. What it proves: the tool is in real use and someone is being hired to run it. How it lies: a template job post copied from an old role, or a "nice to have" mistaken for a requirement. Window: while the role is open. Source: public job boards and company careers pages.

Review-site presence. A company that has reviewed a tool is using it. What it proves: a named person at the account has hands-on experience with the incumbent. How it lies: a positive review is a footprint signal, not a switching signal - it can mean the opposite. Window: none for presence; the sentiment and date are the timing signals below.

Dissatisfaction signals: proof of unmet need

Dated negative review with switching language. A negative review on a competitor's review page is a confirmed buyer publicly documenting an unmet requirement. The reviewer already has budget, understands the category, and is dissatisfied enough to write it down. The highest-value fields are the free-text "cons" and "switching reasons" fields, which reveal objection patterns, integration gaps, and pricing friction. What it proves: a specific, dated dissatisfaction from a named person. How it lies: a manufactured review spike from a campaign, where reviewer profiles cluster by company size, industry, or region. Window: act within 48 hours or the signal is nearly spent.

Public complaint on a forum. Same mechanism as a review, less structured. What it proves: dissatisfaction, sometimes with a named product failure. How it lies: venting without buying authority, or an unresolved ticket that gets resolved quietly. Window: hours to days.

Timing signals: proof of a buying window

Renewal proximity. Most enterprise SaaS contracts auto-renew unless the customer gives notice within a defined window, often 30 to 90 days before the renewal date, and buying committees typically reassemble about 90 to 120 days before contract end. What it proves: a decision point is forced by the contract, not by mood. How it lies: notice windows actually vary from 15 to 180 days, so a default 90-day assumption misfires. Window: engage 90 to 120 days out.

Champion job change. A former customer or champion moves to a new, ICP-fit account. New decision-makers spend around 70% of their budget in their first 100 days, and C-level decision-makers spend roughly $1M on new solutions within their first 90 days in a role. What it proves: an internal advocate now sits where budget authority is fresh. How it lies: the champion has not yet touched the new stack, so day-1 outreach lands before they can act. Window: first 90 to 100 days, first touch on day 14.

Rapid growth the incumbent cannot support. A company running a rival tool while expanding the team that uses it. What it proves: the account is scaling past what the incumbent supports, which is where budget to switch appears. How it lies: hiring for a different function, or growth that the incumbent handles fine. Window: while the hiring surge is live.

M&A or leadership change. Acquisitions force platform consolidation; new leadership on the buyer's team reopens closed decisions. What it proves: an external shock is reopening vendor choices. How it lies: consolidation can favor the incumbent, not you. Window: the quarter around the event.

What each signal proves and how it misleads

This is the fast lookup. Read "what it lies as" as the false positive you will generate if you act on the signal alone.

The single most useful distinction in this whole reference: an account may use a competitor but have no interest in switching, while another may run a complementary tool and be actively researching alternatives. Combining a technographic footprint with a dissatisfaction or timing signal is what separates "relevant" from "in-market." A tool adopted three years ago on a stable stack is not in-market; one just adopted, or one with a renewal 90 days out, is.

Timing beats sentiment: a dissatisfied account is only reachable inside its renewal notice window.

Sourcing and attributing signals from open evidence

Source signals by layering, and attribute person-level signals only with a lawful basis on file. No single source is complete, and no single intent signal justifies a sales touch.

The layered workflow is the practitioner standard: start with tech detection for the initial list, then cross-reference with job postings and review sites. Set the ICP boundary with firmographic filters first, then apply technographic criteria to rank accounts inside it. Record a source count per detection so you can see when a signal rests on one source, and cross-verify before triggering any automated sequence.

For person-level dissatisfaction, the attribution workflow is: read the reviewer's "cons" and "switching reasons" text, pull their public profile to identify role and company, check whether the account is already in your pipeline, and if not, write one email that day - the review's 48-hour half-life leaves no slack.

The reason so many of these signals go unworked is staffing. In Refolk's index, the US has only 285 professionals with a Competitive Intelligence title and the UK has 25.

Most vendors have no one systematically mining competitor reviews. That thin coverage is the opportunity: the signals exist publicly and largely go untouched.

Finding the people behind these signals - reviewers, champions who just changed jobs, teams hiring for a rival's tool - is the part that usually takes a stack of tools and a scraper. This is where a plain-English search removes the friction the paragraph above describes.

I built Refolk to answer questions like that one without a separate detection tool, a job-board scrape, and a spreadsheet merge. The layering still matters; the manual assembly does not have to.

The lawful-basis floor

Scraping is a separate risk from the lawful basis. Platforms have banned and sued operations using fake profiles and pushed data brokers to remove company pages. Keep person-level data minimized to role, company, and business email, and disclose your source in the first contact. A courts ruling that publicly accessible data is not automatically a computer-misuse violation is not a contract waiver and not a GDPR basis; treat it as narrow.

The procedure: from footprint to a ranked switchable list

Run this end to end to turn a competitor footprint into a ranked, defensible list of accounts ready to switch. Steps 1 and 2 build the base; steps 3 and 4 turn it into a ranking; steps 5 through 7 keep outreach lawful, timed, and honest.

How to weight the stack

Use a fit, engagement, intent model with explicit weights, then apply time-based decay, a minimum-signal threshold, and caps so one loud signal cannot dominate. The concrete rule of thumb: a company running a rival CRM and expanding its sales team outranks one running the same CRM with no hiring activity. Pair a static footprint signal with a dynamic growth signal, because that combination is where budget authority to switch actually appears.

How this goes wrong: failure modes and false positives

Most wasted displacement effort comes from a handful of predictable errors. Each one has a check that costs less than the wasted outreach it prevents.

"Uses competitor" mistaken for "ready to leave." The classic false positive is a satisfied, long-tenured user. A tool adopted three years ago on a stable stack is not in-market. Check: require a second signal - renewal proximity, a dropped-tool event, or a dated complaint - before any touch.

Single-source tech detection is stale or wrong. A script tag left behind from a tool already removed reads as a live install. Check: use a source count or independent-confirmation field and cross-reference with job postings, which reflect current hiring intent rather than a cached page.

Manufactured review spikes read as mass dissatisfaction. A sudden burst of reviews may be a campaign, not real churn signal. Check: look for reviewer profiles clustering by company size, industry, or region. Be aware that moderation asymmetry across review platforms can hide real negatives as easily as it surfaces fake positives, and that the same moderation dynamics now span several formerly independent review sites.

Renewal-window guesses. Notice windows vary from 15 to 180 days, so assuming a flat 90 misfires. Check: read the account's actual contract signals rather than applying a default across the list.

Champion-move outreach sent on day 1. The champion has not yet touched the new stack, so a day-1 pitch lands before they have any authority over it. Check: schedule the first touchpoint for day 14, not day 1.

Person-level attribution without a lawful basis. Scraping full profiles or personal contacts is a contract and enforcement risk, separate from the GDPR basis. Check: confirm the LIA exists, data is minimized to role, company, and business email, and the source is disclosed.

Treating a single intent signal as buying intent. No one intent signal is reliable enough to justify a sales touch; a single homepage visit is not a buying signal. Check: enforce the two-to-three signal minimum at the scoring stage, not the rep's discretion.

Reading the published win-rate and cycle-time claims

Do not benchmark against public displacement statistics. The claims conflict badly and none comes from a controlled study, so treat them as direction-of-travel opinions, not numbers you can plan against.

One source says displacement deals close 40% below greenfield but deliver about 3x lifetime value; another says teams doing it well see win rates 35% higher and cycles 20 to 30% shorter; a third says to plan for cycles 30 to 50% longer. These are uncontrolled vendor claims with incompatible methodology. The honest read: displacement can carry higher value per deal, but whether it is faster or slower than net-new is not established in public data. Measure your own lift instead - meeting rate, stage conversion, cycle time, and win rate on displacement plays versus greenfield - and trust that read over any published figure.

Keeping the reference current

A displacement library decays because signals decay. Re-run the validation monthly and retire any signal that has stopped predicting meetings.

The mechanics to re-check rather than memorize: notice windows change per contract, so verify the actual window instead of a default; review-site moderation and ownership shift, so confirm which platforms your competitor's negative reviews actually live on; and scraping enforcement moves, so keep the LIA and data-minimization practice current with what platforms and regulators are actually doing. The static footprint counts move slowly, but the timing and dissatisfaction signals move in hours.

Worked this way, the switchable list is not a static export. It is a queue that reprioritizes as renewal clocks tick, champions move, and complaints get posted, and it stays honest because false positives get reviewed out every month rather than accumulating into noise.