Champion tracking stack — warm outbound on job-change signal instead of cold ICP lists

Every outbound stack on ooligo starts the same way: define an ICP, buy or build a list, sequence it. This one starts from the opposite end. The list already exists — it is your CRM’s contact history, and the signal is that those people keep moving.

UserGems’ own benchmark report, drawn from 2.28 million opportunities across 350+ companies, puts deal cycles 66% longer than 2020 and says it now takes 1.5x more prospects to create one opportunity. Treat those as vendor-published research rather than independent findings, but the direction matches what every SDR manager reports: cold lists convert worse every year, and the cost of the list did not fall to match.

A champion-tracking motion does not fix cold conversion. It routes around it. The shape: UserGems watches the contact base and detects the moves, Clay turns a detected move into a reason to write, Outreach delivers and sequences it, and Salesforce holds the deal history that makes the whole thing worth more than a LinkedIn alert.

How the pieces fit

• UserGems is the signal layer, and the reason it is not just an alert feed is the history it joins. It tracks job changes, promotions, new hires, funding, M&A and tech-stack shifts across the contacts already in your CRM — the vendor markets 30+ native signals. Core includes the Data & Enrichment Agent, the Account & Contact Scoring Agent, Automated Campaigns, and an AI co-pilot Chrome extension. The Writing Agent is an add-on at Core and included from Advanced, which also carries UserGems MCP for building campaigns from Claude or ChatGPT. Credits are the metering unit: Core’s 3.6M credits buy 36,000 data actions or 3.6M AI actions, so the two workloads draw on the same pool at very different rates. On CRM it connects to Salesforce and HubSpot; on delivery, to Outreach, Salesloft and Gong Engage.

• Clay is the research layer that decides whether the email is worth sending. A job change is a timing signal, not a message. Clay takes the new account, runs Claygent against it, and writes back the two or three facts that make the opening line specific — what the new company just raised, what it is hiring for, what changed in its stack. One tier fact decides your plan: job-change signals start at Launch, but CRM integrations, auto-sync and CRM enrichment are Growth-only. A Launch seat cannot read or write the object this stack runs on.

• Outreach is delivery, and its packaging changed shape in 2026. The four packages are Amplify Essentials, Core, Plus and Pro, carrying 10,000 / 25,000 / 50,000 / 100,000 AI credits respectively. Pricing is seat-based plus consumption-based on those credits, so headcount is no longer the only line that moves. Outreach MCP (Server and Client) is listed in every package, which matters here: the same signal base is addressable from Claude on both the UserGems side and the sequencer side.

• Salesforce is the object model, and one edition boundary is load-bearing. Sales Cloud runs Starter Suite at $25/user/month, Pro Suite at $100, Enterprise at $175 (annual billing). Enterprise is the first edition that includes the web API. UserGems writes contacts and signal fields back, Outreach syncs enrollments and activity, and both need that API. If your CRM is on Pro Suite, the stack’s real entry price is a Salesforce upgrade, not a UserGems invoice.

Named handoffs

1. A contact in Salesforce changes employer → UserGems detects the move, creates the contact at the new company, and stamps the prior relationship on it — closed-won at the old account, closed-lost, or champion on an open deal. That stamp is the entire asset. Without it you have a job-change alert, which LinkedIn gives away.
2. The Scoring Agent re-scores accounts and contacts → records above threshold enter an Automated Campaign and are pushed into Outreach as sequence enrollments. Scoring runs on a schedule, so the play is a queue that refills itself rather than a quarterly list pull.
3. Before the first send, Clay enriches the new account row and writes the personalization fields back to the CRM — where the Outreach template merges them. Order matters: enrich on the account, not the person. You already know the person.
4. A champion leaves a current customer → the same detection routes the other way: an at-risk task on the renewal owner and a replacement-champion play, not an outbound sequence. Same signal, opposite motion — see customer-success-expansion-stack and champion-builder.
5. A reply lands in Outreach → the opportunity is created in Salesforce carrying the source signal. Stamp it. If job-change pipeline is not separable from cold pipeline in reporting, you cannot defend the line item at renewal, and this stack’s whole cost case is that its conversion beats cold.

Why this combination

Three of the four tools here can claim “job-change signals,” and Clay’s costs roughly one-fifteenth of UserGems’. So the load-bearing question is what the extra spend buys.

It buys the join. The play that converts is not “this person changed jobs” — it is “this person signed a $180K contract with us at their last company, left eleven months later, and has been in the new seat for six weeks.” That sentence needs the person’s move joined to your own closed-won and closed-lost history, refreshed continuously, across a contact base nobody is curating by hand. Clay is the better tool for a list you are building today; UserGems is the better tool for the base you already own. Buying UserGems without CRM history is buying the join with nothing to join.

The second reason is that the signal has two directions and one owner. The champion who leaves a customer and the champion who lands at a prospect are the same detection event; routing one to an SDR sequence and the other to a renewal task is a configuration, not a second product. Teams that solve only the outbound half rebuild the churn half a year later.

Cost reality

UserGems publishes list pricing, which is rarer than it should be in this category:

• Core — $3,000/month ($36K/year), 3.6M credits, 1x revenue guarantee, $3,000 one-time implementation
• Advanced — $6,250/month ($75K/year), 88,000 data actions or 7.5M AI actions, Writing Agent and UserGems MCP included, 2x guarantee, $5,000 implementation
• Elite — $12,500/month ($150K/year), 214,000 data actions or 15M AI actions, every agent included, 3x guarantee, $10,000 implementation, end users capped at 20
• Retroactive job-change leads — $3,000 one-time at every tier. Buy it once. It backfills the moves that already happened in your base, and it is the only line here that can pay back inside the first quarter.

List is not what buyers pay. Vendr’s published benchmark puts the median negotiated UserGems contract at $30,250 across 172 purchases, ranging $13,275 to $69,569, at roughly 22% average savings — so Core’s $36K list is a starting number, and the median buyer lands under it.

The rest of the stack:

• Clay Growth — $446/month billed annually: $185 for 480,000 actions a year plus $261 for 72,000 data credits. About $5,350/year.
• Outreach — no published price. Vendr reports a median of $45,600/year across 911 purchases, range $8,500 to $213,832, average savings 12.44%.
• Salesforce Enterprise — $175/user/month annual.

Two realistic totals. If Outreach and Salesforce Enterprise are already deployed — the common case — the incremental spend is ≈ $41,400/year plus $3,000 implementation, and $3,000 more for the backfill. Building all four from zero for 15 sellers: add the Outreach median and 15 Salesforce Enterprise seats and you are at ≈ $118,000/year.

Hidden costs the invoice does not show: the contact base has to be clean before any of this fires, because a signal computed on stale records is a wrong-company email to a former buyer (crm-hygiene); one named owner for the signal-to-play mapping, which is a standing job rather than a launch task; and the revenue guarantee measures closed-won, so the clock that decides whether you get a refund runs the length of your longest deal cycle.

Match rules

Right pick: $20M-$500M ARR B2B, deal cycles over 60 days, multi-stakeholder deals with a named champion, at least two to three years of closed-won and closed-lost contact history in the CRM, a sequencer already deployed, and an ACV where $36K is one or two deals rather than a quarter’s target.

Wrong pick: fewer than two years of CRM history, because the asset is the history and it cannot be bought; a self-serve or transactional motion under roughly $10K ACV, where nobody is a champion; a contact base smaller than about 10,000 records, which is the floor of Vendr’s smallest deployment band; or no sequencer, in which case buy that first and come back.

Also wrong if your CRM was migrated recently and closed-lost contacts did not survive the move. Check that specific object before you take a demo — it is the most common way this stack underperforms in month three.

Variations, and the rule for each swap

• HubSpot instead of Salesforce. Both UserGems and Outreach list HubSpot on equal footing. Rule: swap freely on brand — the thing that decides the outcome is whether closed-lost contact history survived your last migration, not which CRM holds it. There is no API-edition gate to plan around either, which makes HubSpot the cheaper entry for teams not already on Salesforce Enterprise.
• Clay alone, no UserGems. Clay carries job-change signals from Launch upward at a fraction of the cost. Rule: take this when the play is “new hires and movers at target accounts” — a list you define — and not “our past buyers, wherever they went.” Clay works the rows you hand it; it does not continuously watch a 50,000-contact base with deal history attached.
• Salesloft or Gong Engage instead of Outreach. UserGems supports all three. Rule: pick by which one already holds your sequences and reporting; there is no signal-side reason to prefer one. See outreach-vs-salesloft.
• UserGems Advanced instead of Core. Rule: the trigger is the Writing Agent, not volume. It is an unpublished-price add-on at Core and included at Advanced along with UserGems MCP. If you want campaigns drafted against your own signal base from Claude, that is the tier — a $39K/year step for two features, so name them before you agree to it.
• Do not make Common Room load-bearing here. Zoom closed its acquisition in July 2026 and the roadmap is now someone else’s. If community and social signal is the actual requirement, read alternatives-to-common-room before wiring it into a stack you plan to keep for three years.

What this stack does not replace

• An ICP or an account list. Champion tracking works the base you already have. New logos with no prior relationship need ai-outbound-prospecting-stack.
• Intent data. Nothing here tells you an unknown account is in-market. Contact-level intent and web de-anonymization are UserGems add-ons at Core and Advanced, priced on request — see intent-data and 6sense.
• A dial floor. No parallel dialing, no connect-rate management. That is outbound-calling-stack.
• Cold-email infrastructure. Secondary domains, warmup and inbox rotation are a separate build with separate economics: cold-email-infrastructure-stack.
• A forecast. This produces pipeline with a source stamp, not a number that survives a board meeting — revenue-forecasting-stack.
• The judgment call about who to contact. A former buyer who churned angrily is a signal too, and the scoring agent does not know the story. Someone has to read the account before the sequence fires.

If you are deciding whether signal-based outbound belongs in your motion at all before you price any of this, start with signal-based-selling and signal-orchestration.