A Lead Source Stops Performing, a Decision Tree
Why this matters
A channel that used to fill the schedule slows down, and the instinct is to either panic and cut it or shrug and blame "the algorithm." Both are guesses. A lead source that stops performing has one of a handful of specific, diagnosable causes, and the fix is completely different depending on which one it is. Cutting a channel that just needs a fresh ad or a rewritten listing wastes a working relationship. Doubling down on a channel that has genuinely gone stale wastes the next quarter's budget. Work the tree before you touch the spend.
Start here: define "stopped performing" in numbers, not a feeling
Before diagnosing anything, confirm what actually changed. Pull the same measure you use to judge every channel (leads per period, cost per lead, or leads converting to booked jobs) and compare the recent stretch against a fair prior window, not just last month against a record best month.
- If the drop lines up with a seasonal pattern you have seen before, this is not a channel failure, it is timing. Compare against the same period last year, not the peak month before it.
- If the drop is real and outside your normal seasonal swing, continue through the branches below in order.
Branch 1: did volume drop, or did quality drop?
These look identical on a revenue chart and have opposite fixes.
- If the number of leads coming in is genuinely lower, the problem is upstream: visibility, budget, or a technical break. Go to Branch 2.
- If leads are still arriving at the same rate but fewer convert to booked jobs, or the jobs that close are smaller or worse-fit, the problem is targeting or messaging, not visibility. Go to Branch 3.
Checking this first saves you from the common mistake of raising spend on a channel whose real problem is that it is now attracting the wrong customer.
Branch 2: volume dropped, check for a break before assuming decay
A sudden or steep volume drop is more often a broken mechanism than a genuinely exhausted channel.
- Confirm the listing, ad, or profile is actually still live and approved. A paused campaign, an expired card on file, a suspended account, or a listing pulled for a policy violation looks identical to "stopped working" from the outside, and is a five-minute fix once found.
- Confirm your own intake is not the leak. A missed call, a broken web form, or a full voicemail box turns a real lead into a no-show in your data, and the channel gets blamed for a failure that happened after the lead already arrived.
- Check whether a competitor materially changed position (a new entrant with aggressive spend, a rival who just fixed their own weak channel). A channel is not decaying, it is being outcompeted for the same limited attention.
- If none of these explain it, the channel itself may genuinely be saturated or declining in reach. Continue to the fix in the next section.
Branch 3: quality dropped, check targeting and message before blaming the channel
If leads keep flowing but the fit gets worse, look at what you are asking the channel to find, not the channel itself.
- Check whether the audience or geographic targeting drifted or was never tightened. A channel that once matched your ideal customer can quietly widen its reach over time, or a setting change on the platform's side can loosen who sees your ad.
- Check whether your offer or message went stale. The exact wording, image, or promotion that worked a year ago can lose its pull once the audience has seen it repeatedly, even though the underlying channel is fine.
- Check whether your own service area or ideal-customer profile shifted (you raised prices, added a premium service line, stopped taking a job type) without updating what the ad or listing promises. A mismatch between what you advertise and what you actually want to sell shows up as declining lead quality, not declining volume.
What to do once you know the cause
| Diagnosis | Action |
|---|---|
| Seasonal pattern | Hold steady, compare against the same period last year |
| Listing or campaign paused, suspended, or broken | Fix the mechanism first, before judging performance |
| Your own intake is losing the lead after it arrives | Fix intake, the channel is not the problem |
| Genuine new competition for the same attention | Refresh creative, tighten targeting, or accept a smaller share |
| Channel reach or attention has genuinely declined | Refresh the offer once; if that does not recover it, retire the channel |
| Targeting or audience drifted wide | Re-tighten the geographic and audience settings |
| Message or offer went stale | Rewrite the ad or listing copy before cutting spend |
| Your service offering shifted without updating the ad | Update what you promise to match what you actually sell |
Give a real fix one fair test before retiring the channel
A channel that has genuinely gone stale deserves one deliberate refresh, not an indefinite decline, before you write it off: a rewritten headline, tightened targeting, and a genuinely different offer, tested for a fair window against your own historical baseline. If a proper refresh does not recover performance within that window, that is your answer, and the money is better spent testing a different channel entirely. What you should not do is keep the same stale creative running while spend quietly increases in an attempt to force the same result.
The order that saves you money
Confirm the drop is real, then check whether it is volume or quality, then check for a broken mechanism before assuming decay, then check targeting and message before blaming the platform. Most "the channel stopped working" panics resolve in the first two steps. The channels that truly need retiring are the minority, and you only get there by ruling out the cheaper, faster explanations first.
References
- Federal Trade Commission, guidance on online advertising performance claims and disclosures
- U.S. Small Business Administration, guidance on evaluating marketing channel effectiveness
- See related: Tracking Lead Source So You Know What Actually Works; Measuring Marketing ROI Without a Marketing Degree