How do I tell if Local Services Ads is adding leads or just taking them from my map listing?
Total lead count cannot answer this, because the LSA unit sits above the map pack and intercepts demand that might have reached you anyway. The only honest test is a holdout: pause the channel in one comparable market or for a defined period, and watch total booked jobs across every source, not LSA's own numbers. Expect noise, and run it long enough to survive seasonality.
Why the platform's numbers cannot settle it
Every ad platform reports the conversions it touched. If a customer would have scrolled past the ads and called from your map listing, the ad still gets to claim the lead when they tap it instead. That is not deception, it is the limit of what a platform can observe about counterfactuals.
So the question is not what LSA reports. It is whether your total booked work goes down when LSA stops. Only your own system knows that, which is why incrementality is a revenue intelligence question rather than an ad reporting question.
Design a holdout you can actually read
A useful test has to beat the noise floor of a seasonal, weather-driven business.
- Geographic holdout. Pause in one market and run in a comparable one. Cleanest design, available only to multi-market operators.
- Time-based holdout. Pause for a defined stretch and compare against the equivalent stretch, ideally against the same period last year rather than last month.
- Measure total, not channel. The outcome variable is booked and completed jobs from all sources combined. If organic and paid search absorb the loss, LSA was less incremental than it looked.
- Change nothing else. No budget shifts, no new promotions, no service area edits during the window.
What partial substitution looks like
The realistic result is rarely all or nothing. Total volume dips less than LSA's lead count, because some demand reroutes to your other listings and some does not. That difference is the incremental share, and it is the number worth knowing.
Watch the mix too. Substituted demand often arrives through slower paths with longer response times, so total volume can hold while booked revenue slips. Comparing sources on completed revenue rather than lead count is the only way to see that, and it needs your ad data joined to your job data through a real integration.
Accept the limits of the answer
A holdout gives you an estimate for one market, one season, one competitive environment. It is not a permanent constant. Competitors enter, Google changes the layout, your review profile shifts, and the number moves.
Re-run it occasionally rather than treating one result as settled. And be honest about the cost: a holdout means deliberately forgoing leads for a period. Some owners will decide that is not worth it, which is a legitimate choice as long as everyone understands that the alternative is not knowing.
Topics: LSA · incrementality · testing · attribution
Have a version of this question about your own business?
The useful answer usually depends on which systems you run and how they're connected. That's a conversation, not a blog post.