Tightly Themed Ad Groups: Anatomy of a Blinds Google Ads Rebuild
The fastest way to waste a window treatment ad budget is one big campaign targeting 'blinds'. When Nova rebuilt the Google Ads account for a seven-figure Indiana window treatment company, the core surgery was structural: the old broad account was torn down and rebuilt as 12 tightly themed ad groups each one a specific service ('custom plantation shutters', 'motorized blinds installer') with its own keywords, its own ad copy, and its own landing experience. Ninety days later: 39% more leads, cost per lead down from $178 to $112 (−37%), lead quality up 86% (4.2 to 7.8), on a budget that FELL from $8,000 to $7,000 a month. This is the anatomy of that rebuild: the full numbers live on the case study page and the template any blinds company can apply to its own account.
Why broad campaigns bleed in this trade
'Blinds' is not one search: it's a big-box shopper hunting $40 miniblinds, a renter measuring for cordless paper shades, and a homeowner ready to spend four figures on motorized somfy-driven rollers, all typing the same word. A catch-all campaign pays for every one of them, shows each the same generic ad, and lands them all on the same homepage. The Indiana account's 'before' was exactly that: broad keywords attracting tire-kickers, climbing cost per lead, and an agency whose answer was to request more budget. Tight theming inverts the economics twice over: relevance cuts the click price (the ad matching the exact search earns better quality scores and cheaper auctions), and specificity pre-qualifies the clicker someone who searched 'custom plantation shutters' has already told you their budget tier.
The anatomy of a tight ad group
- One service, narrowly defined: not 'shades' but 'motorized shades'; not 'blinds' but 'blinds installation'. The Indiana rebuild's twelve groups each owned one such territory: the high-intent services the company actually wanted to sell more of.
- Keywords that all mean that one thing: a handful of close variants, not hundreds of loosely related terms. If a keyword could belong to two groups, it's defined too broadly.
- Ad copy that repeats the searcher's own words: the 'plantation shutters' searcher sees plantation shutters in the headline: the relevance loop that big catch-alls can never close.
- A landing page that keeps the promise: the shutters ad lands on the shutters page, product photos, the consultation CTA, never the homepage. Every mismatch between search, ad, and page is paid for in both auction price and bounce.
- Negatives fencing each group: the group-level exclusions that keep 'motorized' searches out of the manual-blinds group and big-box brand hunters out of everything.
Choosing your twelve (or eight, or fifteen)
The right group list is your service menu weighted by ticket and intent: the premium products that carry the business (shutters, motorization), the volume categories (rollers, faux wood, verticals), the service-shaped searches ('blinds installer near me' behaves differently from product searches, the Indiana account paired this with Local Service Ads for exactly that traffic), and the local modifiers your market actually types. Start from your quote history, the jobs you close profitably name the ad groups you should own, and let the structure mirror the business you want more of, not the taxonomy of a supplier catalogue.
What the rebuild pairs with (the rest of the 90 days)
Structure was the spine, not the whole body: the Indiana engagement added Google Local Service Ads with the Google Guaranteed badge for the near-term service searcher, Meta lookalike audiences built from the customer list with professional video creative, and, the loop that made quality a managed number, call tracking with CRM integration, so every ad group's leads were scored against closed jobs. That measurement layer is why the account could optimize toward its 7.8/10 lead quality instead of raw volume, and why spend could fall while results rose: waste, once visible per ad group, is deletable.
Reading your own account for the disease
- Open your campaigns: fewer than five ad groups for a full-service blinds company almost always means catch-all structure.
- Check keywords per group: dozens of loosely related terms in one group is broad-match sprawl wearing a structure costume.
- Follow three ads to their landing pages: if they all land on the homepage, relevance is leaking at the last step.
- Pull the search-terms report: count the big-box, DIY, and $40-blind searches you paid for last month, that's the rebuild's budget, already being spent.
The takeaway
The Indiana numbers, more leads, better leads, lower cost, smaller budget, came from making every layer of the account agree about one specific buyer at a time. That's all tight theming is: twelve small campaigns that each do one thing honestly, instead of one big campaign that does everything vaguely. The before-and-after, metric by metric, is on the window treatment case study and the account audit that started it is the same one we offer any blinds company, free.