Here's a number worth sitting with: a service-business site sending ad traffic to a general homepage instead of a dedicated landing page sits in the range of cross-industry conversion medians of about 2 to 3% (WordStream via LanderLab). You paid for every one of those clicks. The rest looked, left, and by default you never hear from them again.
Retargeting is the fix for that "again." It's part of the paid acquisition cluster alongside the Google Ads playbook: where that pillar covers building the campaign that earns the first visit, this one is about not wasting it.
What retargeting actually is
Google's own help page describes it simply: retargeting (Google now calls it your data segments, formerly remarketing) lets you show ads to people who previously interacted with your business, as they browse other sites, use apps, or search again. You're not buying a new audience. You're buying a second impression on an audience that already raised its hand once.
That distinction is the whole value proposition. A cold prospect has never heard of you. A retargeted visitor already looked at your services, maybe your pricing, maybe a specific project photo, and left for some ordinary reason: they got interrupted, they wanted to compare, they weren't ready yet. Retargeting doesn't create demand from nothing. It recaptures demand you already paid to generate.
You already spent the money to earn that first visit. Retargeting is how you stop leaving the second one on the table.
Why it converts cheaper than everything else
Cold campaigns have to do two jobs at once: get attention and earn trust, from someone who's never seen you. Retargeting only has to do one: remind someone who already trusts you enough to have visited. That's why it consistently outperforms cold traffic on cost. WordStream says retargeting done right can lift engagement by 400% and sales by 50%, though it does not publish the data behind the claim (WordStream sells ad management). The audience is doing half the work for you before the ad even shows.
That's also why retargeting is a poor place to start and a great place to add on. It has nothing to recapture until your Search, LSA, or Meta campaigns have already sent people to the site. Build the front door first; retargeting is what you install once traffic is coming through it.
Where it fits: after capture, before the decision
If Google captures demand and Meta creates it, retargeting is the layer that sits between both of them and the sale: it holds onto whoever either channel found, for as long as their decision takes to make. A homeowner who Googled "roof replacement cost," read two pages, and left hasn't decided against you. They've decided not yet. Retargeting is what's still in front of them when "not yet" becomes "now," instead of a competitor's ad filling that gap because yours went quiet.
This matters most for the jobs that take real consideration: remodels, roofing, HVAC replacements, legal and financial services, anything in the thousands of dollars where nobody decides on the first visit. It matters far less for a burst pipe at 2 a.m., where there's no multi-week window to catch someone in. Match the tool to the buying behavior, the same logic that decides Google versus Meta.
How to set it up so it works instead of annoys
Four things separate a retargeting campaign that pays for itself from one that just burns a smaller amount of money slowly:
- Segment by intent, not just "visited." Someone who read your pricing page is closer to buying than someone who bounced off the homepage in four seconds. Build separate audiences and messages for each.
- Exclude people who already converted. Nothing wastes budget, or trust, like showing a "get a free quote" ad to someone who already booked. Feed your CRM data back in and cut them out.
- Cap frequency. Somewhere past 5 to 7 impressions a week, more exposure stops persuading and starts irritating. Set a frequency cap; don't let the platform decide for you.
- Give it a window, then judge it. A 30- to 90-day lookback covers most service-business decision cycles. Judge performance on cost per qualified lead, not clicks, and let it run long enough for that window to actually close.
Where this fits in the system
Retargeting doesn't replace anything; it's the connective tissue between the channels that find people and the conversion infrastructure that closes them. One of our clients, Junk Control, built a durable six-figure paid channel in a competitive market by treating every stage of that funnel (from first click to remarketed follow-up) as one connected system rather than a series of one-off campaigns. On the Growth Engine, remarketing sits inside standard managed Google Ads SOP (not a separate unnamed product); the Blueprint maps the gaps, and the Engine runs the campaigns. See /pricing/.
If you're running paid traffic and letting most of it walk away for good, that's usually a structure problem, not a budget problem. It's exactly the kind of gap the Growth Blueprint maps.