Every market has a roofer who can't stand seeing a competitor's ad above their own. They raise the bid, climb back to the top, check again the next morning, and raise it again. By the end of the month they own the first position on almost every roofing search in town, and the invoice from Google looks like a down payment on a truck.
Being first isn't the goal. Booking roofs at a price you can afford is the goal, and those two only overlap on some searches. This chapter is about finding which ones.
Hand-Set Bids Stop a Small Repair Search From Costing What a Full Replacement Does
Isolation, the exact-match roofing campaign from Chapter 6, runs on bids you type in yourself, one per ad group. That's not nostalgia for manual work. It's the only way to tell Google that a replacement search and a small-repair search are worth different money to you.
Go back to the job math from Chapter 3. A full replacement might be a five-figure job. Fixing a few blown-off shingles might be a few hundred dollars. If both ad groups carry the same bid, you're either overpaying for repair clicks or underbidding on the searches that pay for your crews. Hand-set bids let you price each ad group to what its jobs are worth, and adjust as the leads come in.
A simple way to start: set the replacement and hail groups highest, the repair group lower, and anything that tends to produce small service calls lowest. Then let a month of real leads tell you where you were wrong.
Keyword Planner's top-of-page range gives you a starting point for each group. Opening near the upper end on the replacement and hail groups, as Chapter 6 explained, gets a new account seen. The repair group can open nearer the middle. What you're really setting is a ceiling: the most you'll pay for one homeowner's click on that kind of job. Write those ceilings down. When the leads start coming in, you'll compare them against what each group actually paid per booked inspection, and the gap between the two tells you which way to move.
A Roofing Keyword Worth Winning Books Inspections, Not Just Clicks
Our method sets a clear target for the money keywords: show at the top of the page, above the free results, on roughly 75% to 90% of the searches where you're eligible. Google reports this as top impression share, and a companion number, absolute top impression share, tells you how often you held the very first spot.
The trick is the word "money." Money keywords are the ones that book inspections, which is why Chapter 1's tracking has to come first. Pull up the keywords that produced booked inspections and signed contracts over the last month or two. Those deserve top-of-page bids. The rest, the searches that generate clicks and the occasional price shopper, can live lower on the page at a lower cost. A click never climbed a ladder. An inspection does, and that's the only thing worth paying top dollar for.
On a hail week, the list of keywords worth winning can change overnight. That's fine. Hand-set bids can change overnight too.
Lost Impression Share Tells You Whether Budget or Rank Is Costing You Roofs
When your ad doesn't show, Google tells you why, and there are two different reasons with two different fixes.
- Lost to budget: Google's numbers count the share of searches where your ad didn't appear because the campaign's budget had run out. The fix is more budget, or fewer keywords so the budget you have covers the ones that matter.
- Lost to rank: here the ad was eligible, but its Ad Rank wasn't high enough to show. The fix is a better ad, a faster and more focused landing page (Chapter 4 covers both), or a higher bid on the searches that justify one.
Both are columns you can add in Google Ads: Search lost IS (budget) and Search lost IS (rank). Google reports the budget number for the whole campaign only, so if one campaign carries both your replacement and your repair searches, you can't tell which one ran dry. That's one more reason the money searches deserve room to breathe. Check both numbers weekly, and start with the campaigns that book roofs.
Mixing them up is expensive. A roofer losing to rank who raises the budget just runs out of money on the same bad positions. A roofer losing to budget who rewrites every ad fixes nothing that was broken.
Lead Sellers Bid on the Same Roofing Searches You Do
Here's something roofers often don't see until they look. While researching our example market, Minneapolis-St. Paul, we checked Google's Ads Transparency Center: Angi, HomeAdvisor, Modernize, Thumbtack and HomeGuide were all running Search ads in the three months before we looked, alongside national roofing and exterior brands. The Transparency Center doesn't show which cities an ad targets, but the same lead sellers also showed up in the free results we pulled for Twin Cities roofing searches.
Those advertisers aren't roofers. Lead sellers pass the homeowner's request along to contractors, and their business model is different from yours. That means their bid on "roofing companies near me" isn't set by what one of your roofs is worth. It's set by what they can sell the lead for. You're bidding against a middleman who gets paid whether or not you get the job.
Google's auction insights report shows you who you're meeting. Pick a roofing keyword, open the report, and there's the lineup: every advertiser who landed in those auctions with you, plus a few percentages. One says how often a rival turned up when you did. Another says how often they sat higher on the page. A third says how often you beat them.
Our method has a simple rule for this: if an auction gets too expensive, step out of it. You don't have to beat a lead seller on every search. If they're outbidding you on a generic search that rarely books, let them have it and put that money into the searches where your ad, your page and your reviews win the job. Chapter 15 covers lead sellers in more depth.
City-Name Roofing Searches Carried the Highest Bids in Our Example Market
Not every search costs the same, and the pattern isn't always obvious. In our example market's Keyword Planner data, the searches with the city's name in them topped out higher than the same searches without it:
- "Roofing companies near me": $20 to $85. The city version: $27 to $112.
- "Roof repair near me": $26 to $90. The city version: $33 to $101.
- "Roofers near me": $18 to $75. The city version: $18 to $96.
- "Roof replacement": $18 to $100. The city version: $30 to $132, the highest top-end bid in the pull.
These are Keyword Planner estimates built on broad match, so treat them as give or take about 30%. They don't tell you the city searches convert better. They tell you advertisers are paying more for them, which is a reason to check your own numbers before you match their bids. Following the crowd's bid is how a roofer ends up paying tourist prices in their own hometown. Your own market's pattern may be different; the lesson is to look, not to assume.
Broad Match's Target Should Be the Lead Price Exact Match Already Proved
When the Exploration campaign launches, bidding changes hands. Instead of setting each bid yourself, you give Google a target cost per lead and let its automated bidding set bids auction by auction.
Where that target comes from matters. Use what the Isolation campaign actually paid per lead over its last month or two, not the price you'd like to pay. Google's help page describes the target as a running average across many leads rather than a ceiling on each one: some leads will come in above it, some below. One more line on that page matters to roofers: Google may spend as much as double your average daily budget on a busy day. After a hailstorm, that busy day is coming.
A target that's too low doesn't save money. It just starves the campaign, because Google can't find enough homeowners at that price and stops bidding in the auctions that would have booked jobs. Start honest, let the campaign produce, then tighten the target in small steps.
Google will usually suggest a target for you, and on a campaign with conversion history it's built from roughly the last 30 days of average cost per conversion. That's a reasonable sanity check. But if those 30 days included a hail week, the number may be flattering, and a target set on a storm's best week will choke the campaign in a quiet October.
A Call That Books Nothing Shouldn't Set Your Roofing Bids
Every bid in the account is only as good as the conversions it learns from. For roofers, a lot of those conversions are phone calls, and not every call is a homeowner.
Our method includes a regular call quality review. Look at the calls that counted as conversions: how long they lasted, where they came from, what they were about. A roofing account can pick up wrong numbers, vendors selling you something, job seekers asking if you're hiring and existing customers calling about a warranty. Each one that counts as a lead teaches the bidding to find more of the same.
Chapter 1 set a 30-second minimum for calls, which filters out the shortest noise. The review catches the rest. If a pattern of junk calls keeps coming from one campaign or one time of day, adjust the tracking or the targeting before you adjust the bids.
The review doesn't need to be elaborate. Once a week, pull the call log for the calls that counted, sort by campaign, and have whoever answers the phone mark each one: booked, real but didn't book, or junk. Twenty minutes with a coffee. The junk column is the one that matters. If it's growing, the bidding is learning the wrong lesson, and every dollar you add teaches it faster.
Next: Headlines in Homeowners' Words
You now know which searches deserve the top of the page and how to pay for them without draining the budget. The next piece is what the ad actually says once it's there. Chapter 8 pulls homeowners' own words from hundreds of roofing reviews and turns them into headlines.
The full guide is on the roofers guide home, and when a roofing company wants someone else managing the bids, that's what our Google Ads management does.
