FOOTPRINTSEO Training Center
(239) 457-1867
New
  • Module 093
  • 2026-10-22
  • "REX (Matt Diamante / Get-Found school)"
  • "Local SEO / Google Business Profile"

Module 93: GBP Custom Services — The Keyword-Variation Rule That Actually Moves the Local Pack

The Signal — What Changed

A Local Search Forum thread on October 22, 2026 asked the question every local SEO eventually fights about: when you fill out custom services on a Google Business Profile, how many keyword variations are actually worth adding? Is the GBP services list a free keyword buffet where "air conditioner repair," "AC repair," "air conditioning repair," "AC fix," "HVAC repair," and "emergency air conditioner repair" all earn their own slot — or does stacking near-duplicates hurt?

The thread lands the same week Google is testing a layout change for the Local Knowledge Panel (Search Engine Roundtable, Oct 22, 2026). When Google redesigns the local panel, the surface area it gives to services, photos, Posts, and reviews shifts. A services list that was invisible last month can suddenly be front-and-center — and vice versa. That makes services hygiene a live ranking-adjacent surface again, not a set-and-forget checkbox.

The answer from the LocalU / Sterling Sky school (Joy Hawkins' forum) is consistent with years of local-pack testing: services relevance matters, but duplicate stuffing dilutes and can look spammy. One service per real offering, and variations only where customers actually use different words.

REX take: the GBP services field is a shelf, not a keyword dump. Stock one clean label per real service, use the words your customers actually say, and let photos, reviews, and posts on the same profile reinforce the same vocabulary. Breadth without substance reads as spam — to Google and to the human scanning the panel.

Why It Works

Google's local algorithm matches a searcher's query against the categories, services, name, and on-profile signals of a business — then against proximity and prominence. The services field is one of the few profile elements you fully control. But relevance signals don't stack linearly:

  • Real coverage wins. Listing "Emergency AC Repair" as a distinct service can legitimately match a query the generic "AC Repair" entry ranks weaker for — because it's a real service people actually search at 2am.
  • Near-duplicate stuffing doesn't. "AC Repair," "A/C Repair," "Air Conditioner Repair," and "Air Conditioning Repair" are the same service with spelling variants. Google's language understanding already collapses those variants; five entries covering one service just spreads your relevance thin and reads as keyword manipulation.
  • Services reinforce the whole profile. The words you put in services appear in panel surfaces, feed the topical association of the profile, and (per the LSF thread) work best when the website page the profile links to carries the same services vocabulary. Profile and landing page telling one coherent story beats either alone.
  • The panel redesign raises the stakes. With Google testing the Local Knowledge Panel layout (Oct 22), a well-stocked, well-named services list is cheap insurance: whichever way the layout lands, your shelf is stocked.

How to Run It

Step 1 — Inventory real services, not keywords. List every service the business genuinely sells as its own offering. For an HVAC company: AC Repair, AC Installation, Furnace Repair, Furnace Replacement, Heat Pump Service, Ductless Mini-Splits, Maintenance Plans, Emergency HVAC. That's ~8 real services. Then look at your invoices/call logs: which ones do customers book as separate line items? Those are your services entries.

Step 2 — Apply the variation rule. For each service, ask: do customers actually call it something else in real conversations, reviews, or search queries? If call logs show people say "no-heat call" and the service is furnace repair, adding "No-Heat Emergency Furnace Repair" is a justified variation. If the only difference is spelling or word order ("AC Repair" vs "A/C Repair"), do not add a second entry.

Step 3 — Match the landing page vocabulary. Each custom service should map to (or at least be mentioned on) the relevant page of the website the GBP links to. If the profile says "Heat Pump Installation" and the site has a heat-pump page, the signal reinforces. If the site has no heat-pump content, either add a short section to the main service page or expect that service entry to underperform.

Step 4 — Write descriptions like a human. Every service entry supports a short description. Use it: 1–2 sentences with the service, the city, and one concrete detail ("Same-day AC repair across Kennesaw and Cobb County — most repairs completed on the first visit"). No keyword strings, no other business names.

Step 5 — Local service checklist. Cap total services at roughly 8–15 for a single-location contractor. Prioritize: (1) highest-revenue services, (2) emergency variants customers genuinely ask for, (3) seasonal services updated quarterly. Screenshot the services panel before and after — you want a before/after record for the weekly panel-layout log.

Step 6 — Insurance/YMYL version. For regulated verticals (Medicare, ACA, insurance services), keep service/offer labels strictly factual and compliant: "Medicare Advantage Enrollment Help," "ACA Marketplace Quote Consultation," "Supplemental Coverage Review." Never use services text to promise savings, eligibility, or outcomes — the same compliance rule that governs the website governs the profile. Name the service, not the outcome.

Real Example

Client: acrepairdaily.com. Their GBP had 23 custom services, including "AC Repair," "A/C Repair," "Air Conditioner Repair," "Air Conditioner Repair Near Me," "HVAC Repair Near Me," and "Best AC Repair" — five entries for one service plus two keyword-stuffed variants. The panel looked spammy to humans, and the profile's relevance was smeared across duplicates.

The cleanup took 40 minutes:

  • Collapsed the five repair variants into one "AC Repair" entry with a description that naturally contained the variants: "Air conditioning repair in Roswell — from a failing compressor to a system that won't turn on, same-day diagnosis on most calls."
  • Kept "Emergency AC Repair" as a separate entry — call logs proved customers book it as its own thing after-hours.
  • Deleted "Best AC Repair" and "Near Me" entries entirely — those are search phrasing, not services.
  • Ended at 9 clean services, each mapped to a page section on the site, each with a human-written description naming the service city.

Result to watch: cleaner panel presentation for the humans scanning it, tighter topical focus for the algorithm, and a before/after screenshot set for the weekly panel-layout log (relevant immediately given Google's October panel layout test). Expect any movement in pack placement over the following 2–4 weeks, tracked in the Friday ops check.

Checklist

  • [ ] Full inventory of real services from invoices/call logs
  • [ ] Variation test applied: separate entry only for services customers genuinely name differently
  • [ ] No "near me," "best," or city-in-name keyword stuffing entries
  • [ ] Each service maps to website content on the GBP landing page (or a section of it)
  • [ ] Every service has a 1–2 sentence human description with city + concrete detail
  • [ ] Total service count within 8–15 for a single-location contractor
  • [ ] YMYL/insurance profiles: factual service labels, zero outcome promises
  • [ ] Before/after screenshots saved to the weekly panel-layout log

Quiz

Q1: You have "AC Repair," "A/C Repair," and "Air Conditioner Repair" all as separate GBP services. What's the right move?
A) Keep all three — more entries means more keywords
B) Collapse them into one "AC Repair" entry with a description that naturally covers the wording variants
C) Delete the services section entirely
Answer: B — spelling/word-order variants are the same service; one clean entry with natural language in the description preserves relevance without dilution.

Q2: Which of these justifies a separate custom services entry?
A) "Best Roof Repair Near Me" — it's what people type
B) "Emergency Roof Repair" — call logs show after-hours customers book it as its own service
C) "Roof Repair Johns Creek" — city pages help rankings
Answer: B — an emergency variant is a real, separately-demanded service. "Best/Near Me" and city-tagged entries are search phrasing, not services.

Q3: True or false — for an insurance agency's GBP, services should promise outcomes like "Save $500 on Medicare Plans."
A) True — promises drive clicks
B) False — services labels must stay factual and compliant ("Medicare Plan Enrollment Help"); outcome promises violate the compliance rules that govern both the website and the profile
Answer: B — name the service, never the outcome.

Try This Today

Pull up one managed GBP's services list right now and run the 10-minute dedupe: delete every "near me"/"best"/city-stuffed entry, collapse spelling variants into one entry each, and add a real human description to the top three revenue services. Screenshot before and after — and log it in the weekly panel-layout comparison while Google's layout test is live.

Sources