Most local websites get designed in a room where nobody has looked at a single search query. Somebody picks the colors, somebody argues about logo size, and the page list gets copied from the last site. Six weeks later the site looks better and produces the same number of calls.
Design is rarely the cause. The calls stay flat because nobody asked what your customers in Edwardsville, Glen Carbon, and Collinsville type into their phones at 7:40 on a Tuesday morning, or what those phrases are worth. On July 30, 2026 I pulled live demand and live results pages for a set of Metro East trade keywords, and every number below came out of that pull or a study I can name.
What data-driven design decides
Data-driven design means your sitemap, page headlines, and homepage promise get chosen from measured demand in your service area rather than from opinion in a meeting.
Which pages exist. Which services get their own URL. Which towns get a page. Taste usually decides all of that, and a number can decide it instead. Whitespark's 2026 survey of 47 local search experts ranked a dedicated page for each service as the top local organic factor. Start with the pages a small business site needs and let the data extend the list.
How to find what people in your towns are searching
Pull volumes at the metro level from a keyword tool, then confirm each candidate against the live results page before trusting it.
Metro level matters more than most operators expect. The St. Louis DMA covers 31 counties, Madison and St. Clair among them, and pulling at the Illinois level folds in Chicago. Same keywords, same day:
| Keyword | Illinois | St. Louis DMA | Inflation |
|---|---|---|---|
| plumber near me | 33,100 | 8,100 | 4.09x |
| ac repair near me | 6,600 | 1,900 | 3.47x |
| plumber edwardsville il | 140 | 140 | 1.00x |
| hvac belleville il | 110 | 110 | 1.00x |
Terms carrying a town name read identically at both levels, because the modifier locks the geography. Terms without one inflate three to four times. Two free sources sit alongside the tool: Search Console, for queries your site already earns impressions on, and your own phone log.
Why volume alone will point you at the wrong pages
Search volume is a rounded, bucketed, keyword-grouped annual average, and the distribution skews so hard that sorting by it surfaces the terms a local business cannot win.
Ahrefs tested 72,635 keywords against Search Console data and found Keyword Planner overestimated volume 54.28% of the time. In a separate count of its own US database, about 93% of keywords get fewer than ten searches a month. Sort an export by volume and you have sorted a curve containing almost none of your customers.
Zero reported volume is its own trap. "Water heater replacement edwardsville il" returned a null volume at the DMA level, while its live results page carried a map pack and ten organic results, every one a town-specific URL. Competitors had built pages for a term the tool says does not exist. Treat a null town-level term as unknown, not dead.
Intent is the real signal, and the ad market prices it
Intent tells you whether a searcher wants to hire someone or avoid hiring someone, and cost per click is the cleanest available measurement of it.
Read intent off the modifier pattern:
- Urgency words ("emergency", "same day") are the top transactional tier.
- A town name or "near me" is transactional and local.
- Best, top, reviews, cheapest is commercial investigation, and cost or price is money-adjacent but not ready.
- How to, why, DIY is informational, and in the trades it means someone fixing it themselves.
That last line is the correction most sites miss. A how-to query signals a person avoiding you, and the ad market prices the gap: "how to unclog a drain" carried a $4.18 cost per click while "drain cleaning near me" carried $32.97. Hold volume constant and the spread widens. Five terms, all at exactly 70 monthly DMA searches:
| Keyword | Volume | CPC | Intent |
|---|---|---|---|
| furnace filter sizes | 70 | $1.01 | Informational |
| how to unclog a drain | 70 | $4.18 | Informational |
| sewer camera inspection | 70 | $14.80 | Commercial |
| sump pump replacement cost | 70 | $27.65 | Commercial |
| sump pump installation near me | 70 | $37.86 | Transactional |
Identical volume, a 37.5x spread in market value, and only one of the five ends in a truck at somebody's house. The gap widens once AI results enter. Whitespark's May 2025 study found AI Overviews on 92% of informational local queries but 15% of local-intent ones, and every Edwardsville term I pulled returned a map pack with no AI Overview. Google answers the how-to questions itself.
How to score a term before you build a page for it
Score every candidate as reachable volume multiplied by intent price multiplied by winnability, then rank by the product rather than any single input.
Reachable volume is DMA volume discounted to the share of the metro you serve. A plumber working Madison County covers a single-digit slice of a 31-county metro, so the example below discounts by 8%.
Intent price is cost per click, taken directly. Where CPC comes back null, no advertiser wants the term, so treat it as informational however commercial it looks.
Winnability comes off the live results page rather than a difficulty score. Read the map pack review counts first: the pack for "ac repair near me" in Edwardsville showed 181 reviews, 96, and 5,500. A 96-review pack is a season of disciplined review requests away. A 5,500-review regional is a moat. Then count the directory slots in the organic top ten. On "plumber edwardsville il," a Facebook post, Yelp, and the BBB held three of them, each a position a local business page can displace.
Advertiser competition points at the same openings. The Google Ads competition index runs 0 to 1:
| Keyword | Volume | CPC | Competition |
|---|---|---|---|
| drain cleaning near me | 880 | $32.97 | 0.13 |
| water heater repair near me | 720 | $56.09 | 0.15 |
| emergency plumber near me | 1,000 | $65.59 | 0.20 |
| plumber edwardsville il | 140 | $23.13 | 0.98 |
The flagship town term is saturated at 0.98 while the higher-value specific-service terms sit at 0.13 to 0.20. Own the flagship term for credibility, and take the near-term wins from the cluster nobody is bidding on. Keyword gap tools compare your domain to competitor domains, so directory slots never enter the comparison.
Why every page needs a term behind it
A page should exist because a measurable number of people search for the thing that page is about, and for no other reason. A "Services" page listing eleven services in bullets competes for nothing. Eleven pages, each built on a term with real demand, compete for eleven things.
The near-me question resolves itself here too. Searching "ac repair near me" from Edwardsville, eight of ten organic results were town-specific URLs like /ac-repair-edwardsville-il/. Google rewrote the query into the searcher's town, so the 1,900 monthly DMA searches on that phrase get captured by the page targeting "ac repair Edwardsville IL." Build the town page and let Google do the rewrite, which is much of what makes trade business websites work in this market.
A worked example: two plans for an Edwardsville plumber
Given the same business and the same budget, an intent-first page plan is worth roughly 100 times a volume-first plan.
Take a plumber in Edwardsville serving Glen Carbon, Maryville, Troy, and Collinsville. Call 8% of the DMA reachable.
The volume-first plan. A tool export sorted by volume produces the informational cluster: "how to unclog a drain" at 70 a month, "furnace filter sizes" at 70, "why is my ac not cooling" at 20. Total 170 monthly DMA searches, about 14 reachable, at a blended CPC near $2.50: roughly $35 a month of market value. Most of those results pages now carry an AI Overview, so Google answers them before anyone clicks.
The intent-first plan. The transactional cluster: "plumber near me" 8,100, "emergency plumber near me" 1,000, "drain cleaning near me" 880, "water heater repair near me" 720, "sump pump installation near me" 70. Total 10,770 monthly DMA searches, about 860 reachable, plus the geo-locked town terms that take no discount. Blended CPC runs near $40. Assume a 10% capture: roughly 86 clicks a month, near $3,400 of equivalent ad value, and at an assumed 8% conversion rate about seven leads.
Roughly $3,400 against roughly $35. The page set that delivers it:
- /drain-cleaning-edwardsville-il/
- /water-heater-repair-edwardsville-il/
- /emergency-plumber-edwardsville-il/
- /sump-pump-installation-edwardsville-il/
- /service-areas/glen-carbon/, /maryville/, /troy/, /collinsville/
That is a sitemap derived from measured demand, longer than most Metro East trade sites carry. Preparing for a redesign starts with the page list, and the page list starts here.
What this does to your homepage headline
Your homepage headline should state your highest-value transactional term as a promise, in the words people search, with your towns named.
Most local homepages open with a slogan: "Quality you can trust since 1987." Nobody searches that. The searcher who typed "emergency plumber near me" needs to see in a second that they landed on one, in their town. "Emergency plumbing in Edwardsville, Glen Carbon, and Collinsville. Same-day service, licensed and insured." That headline matches the query and confirms the geography.
The same logic runs down the page. Service blocks follow value order, so drain cleaning and water heater repair sit above the low-CPC filler. The phone number stays visible at every scroll position, because the call is the conversion.
None of this makes the site uglier. Taste still decides typography, photography, and spacing. Data decides what the pages are and what they say.
Where to start
Pick your five biggest services and your five real towns, pull volume and cost per click for each pair, then read the live results page for the ten that score highest. That is an afternoon of work.
The list you end up with is your sitemap, settled before anyone opens a design tool. If you would rather hand the research and the build to someone else, LocalCraft Digital does both together, starting with a free build.