Database reactivation, explained: what it is and how to run it

Four words people use interchangeably, and shouldn't

Most confusion about DBR comes from treating one pile of records as one audience. These four categories carry different evidence and warrant different messages.

• Dormant customer — someone who bought before and hasn't returned within your normal repeat cycle. The evidence is a completed transaction and a working relationship. This is the strongest group, because you have delivery history to reference
• Aged lead — an inquiry that never reached a priced proposal. You know they raised a hand; you don't know what they wanted or what it was worth. The weakest evidence of the four
• Unsold estimate — a specific, priced proposal that was sent and never became a job. The strongest commercial evidence: a defined scope, a real number, and a decision that was never explicitly made
• Stalled opportunity — anything that moved through part of your process and then stopped mid-flow: a revision requested and never returned, an approval never scheduled, a deposit paid against work never booked. These are usually the highest-conversion rows in any export and the ones most often overlooked, because the record doesn't look closed

Why records go quiet, and why that isn't the same as "no"

A lead can go cold for reasons that expire. Budget timing, a delayed move, a competing household priority, a landlord decision, sticker shock that faded over a season — none of these are rejections, and none of them announce themselves when they resolve. The customer simply doesn't circle back, because circling back requires them to remember your name.

What a status of sent or pending proves is narrower: the opportunity was never closed cleanly. It doesn't prove the work is recoverable. Holding both of those ideas at once is the difference between a realistic campaign and a fantasy about a spreadsheet full of money.

How to segment the export

Once the file is out of the CRM, three cuts produce most of the value. Layering them gives you a small number of meaningfully different groups instead of one undifferentiated blast.

• By value — split at your own average job size. Above it, a human call and a written-by-hand email are worth the time. Below it, the standard sequence carries the load
• By recency — under 90 days can be reopened as a continuation of the same conversation. Roughly 90 days to a year needs a reintroduction. Beyond a year, decide deliberately whether contact is appropriate at all, and say plainly that time has passed if you proceed
• By reachability — rows with no email and no phone are not an audience, they're a data problem. Count them separately so they don't inflate your projections
• By category — dormant customers, aged leads, unsold estimates, and stalled opportunities each get their own message. A past customer should never receive copy written for a stranger

How to measure whether it worked

This is where most DBR efforts fall apart, because open rates are easy to report and prove nothing. The measurements that matter are further down the funnel and take longer to collect.

Track, per segment: replies received, conversations that reached a quote or re-quote, jobs booked, and revenue actually invoiced. Revenue is the only number that survives contact with a bank account. Also track the negative signals — opt-outs, complaints, and hard bounces — because a campaign that books work while burning your sending reputation has borrowed against next quarter.

If your list is large enough, hold back a randomly selected slice and leave it uncontacted. Some of those records would have come back on their own, and without a holdout you have no way to separate what the campaign caused from what would have happened anyway. Compare bookings in the contacted group against the holdout over the same window. It's the difference between attribution and a story.

Set the measurement window generously. A reactivated estimate can take weeks to become a scheduled job, and judging results at day seven will make a working campaign look like a failure.

Five steps, in order

• Export — pull estimates and customers out of your CRM or spreadsheet, keeping the status and date fields intact
• Clean — merge duplicates, normalize phone and email formats, and classify each record as won, lost, or genuinely unresolved
• Segment — apply the value, recency, reachability, and category cuts above
• Sequence — a small number of touches over two to three weeks, written to start a conversation rather than to announce an offer, on channels each contact is eligible to receive
• Prove — measure booked revenue by segment against a holdout, and let that decide what happens next

Agency, DIY, or product

Three ways to run this. An agency will build and operate the campaign for you, typically on a retainer plus a share of what it produces; the practical questions to ask are who owns the data and the sending domain when the engagement ends. Building it yourself in a general marketing tool is entirely possible, and the work that sinks DIY attempts is rarely the sending — it's consent tracking, suppression across channels, and keeping up with replies. A purpose-built product trades flexibility for having those decisions already made.

We're the third option and openly biased about it; the side-by-side comparison lives on our GoHighLevel agencies page. Whichever route you take, start from the measured number rather than an estimate of it. SignalBack's free CSV audit reads your export and reports what's actually in there — that audit is the part of the product that's live today.