Headstone Cleaning
Keyword Field

31,880 searches a month, worth about $43,000 in equivalent ad spend

That is the honest number for everything a cemetery-cleaning business could actually sell against, nationally, after collapsing Google's close-variant duplicates. The naive version of this number, the one you get by summing a keyword export's volume column, is 3.8× larger, and it is wrong.

The collapse rule is deliberately conservative in the other direction too: identical volume/CPC pairs are merged only above 50/mo. Below that, 55 unrelated long-tails all report exactly 10/mo at $0.00, which is Google's rounding floor showing through. Merging on it would under-count the tail as badly as the naive sum over-counts the head.

Two thirds of the discovered pool is not this market

Seeded with ten cemetery-cleaning terms, DataForSEO's keyword_ideas endpoint returned the whole cleaning services and funeral home categories, plus indeed (13.6M/mo) and business name generator. It associates by category, where the seed terms are topical. Relevance had to be decided on two axes at once: is this the same subject, and is this someone who would buy.

Ten clusters, and the money is concentrated in three

Note the inversion between the top two rows. Product/DIY carries more raw searches than core service (49,940 vs 47,730) at less than half the CPC. People researching D/2 Biological Solution outnumber people looking to hire, and they are the same people, earlier. That is the content wedge.

The commercial head, priced

Restoration is the premium line: headstone restoration services costs $10.54 a click against $0.77 for the generic cleaning head. Same customer, later in the decision, worth 13× more per click.

The metros are real but thin. City-name searches don't exist.

DataForSEO Labs is country-only, so metro demand came from Google Ads search_volume at DMA and city codes: the same instrument at every level, so the ratios mean something.

City-level demand, all 355 addressable keywords per city

Top 15 of 30 cities priced. Renton is the outlier worth noticing: only 420 searches a month, but at $4.76 a click it carries more measured value than Everett, Tacoma and Bellevue combined.

May is 5.5× February. Everything else is a rounding error.

Twelve months of history came down with every keyword and had not been looked at. It turns out to be the largest single effect in this dataset, bigger than the gap between the two metros and bigger than any cluster.

No cleaning company owns this SERP. Google does.

Across 110 live SERPs in three geographies, the most common domains are platforms. This niche has never been properly contested.

Every dedicated cleaning operator Google Maps will show, in all 51 states

The competitor table earlier was drawn from three geographies, so it could only ever see firms that rank nationally, in LA or in Seattle. This asked Google Maps once per state per head term, 153 searches in all, and it is the only instrument used here that returns a street address, so a firm's state is read off its own listing rather than inferred. Then every survivor was opened in a browser and looked at, one at a time, because a business name does not say what a business sells: 12 of the 74 that passed the name filter sell something else: marble and terrazzo firms trading on the word restoration, a dry cleaner, an Airbnb cleaning service whose owner is named Graves, a monument catalog, a British firm mis-geocoded into Washington. They are listed at the end, with what each one does. What remains sells the cleaning.

The same 62 operators, geographically

Every listing came back with its own coordinates, so the field can be read as a map instead of an alphabet. Search it, filter it to storefronts or service-area firms, and click any dot for the listing behind it. Circle size is review count, which is the closest thing this market has to a size signal.

States = how many state-level searches the firm surfaced in, a reach signal: 1 means a purely local business, double digits means a service-area operator bidding for the whole country. ~ marks a firm with no published street address, where the state comes from its own telephone area code, or failing that the state it ranks best in.

The 12 that did not survive the screenshots

Every one of these ranks for a headstone-cleaning search and passed a filter built on business names and Google categories. Every one of them sells something else. They are here because the filter that let them through is still the filter, and the next refresh will let something similar through again. Opening the page is what catches it.

120 People Also Ask questions, and one of them is the business

Harvested from the same 110 SERPs. The count is how many separate SERPs each question appeared on, a proxy for how central Google thinks it is.

Five moves, in this order

Ordered by evidence. Each one names the finding it rests on.

Limits

• No client, no domain. This survey was run without a target site, so there is no ranked-keyword baseline, no backlink profile, and no gap analysis against a specific competitor. Pointing it at a domain adds ranked_keywords and competitors_domain for about $0.25.
• Volume is Google's estimate. Search volumes are 12-month averages from Google Ads, rounded hard at the low end; the 10/mo floor covers anything from 1 to 49.
• The value column is a proxy. Volume × CPC is what the clicks would cost to buy. It ranks opportunities; it does not forecast a P&L.
• 15 cities per metro. Bothell, Lynnwood and Shoreline all returned demand once measured; the tail of smaller suburbs is unpriced.
• The directory is what Maps shows. A firm with no Google Business Profile does not exist to this method, and three head terms at 20 results per state is a wide net but not an exhaustive one. Read "15 states with no operator" as "nobody is visible there", which is the thing that matters commercially, not as proof that nobody is trading.
• 20 of 62 operators have no published street address. Their state comes from the firm's own telephone area code, which is good evidence. A business can keep a number after moving. Two firms had neither address nor a geographic area code and are placed by where they rank best. Three of the 62 are listed but not plotted on the map: Google returned no usable coordinate for them, two of them sharing an identical placeholder point in the Pacific.
• Seasonality is national and category-wide. The monthly shape is summed across the whole addressable pool; individual keywords will have their own curves, and the two metros were not broken out separately.