Do long-tail searches convert better on Google Ads?
Not on cost: across 28 home service Google Ads accounts (Aug–Oct 2026), 4-word searches had a median cost-per-conversion index of 113.61 (22 accounts) and 5+ word searches 111.3 (18), against 97.81 for 1–2 word searches (25), where 100 is each account's own average.
- Three-word searches were the one length where most accounts converted better than their own average: 13 of 20. Their cost per conversion index was 99.4.
- Three- and four-word searches cost more per click: 106.12 and 108.73, against 97.87 for 1–2 words. Five-plus word searches were 94.66.
- Junk removal accounts only: 4-word searches came in at 121.58 (17 accounts).
- Searches of 1–2 words were a median 33.87% of each account's clicks (36 accounts).
By Justin Hubbard, who runs Adimize and has owned and operated a junk removal and dumpster rental company for over 12 years
Based on 28 home service Google Ads accounts (per-account index) and 36 accounts (click share), Search campaigns, Aug - Oct 2026
Cost per conversion by search length (100 = the account average)
| Search length | Cost per conversion index | Cost per click index | Conversion rate index | Accounts | Converted better than own average |
|---|---|---|---|---|---|
| 1–2 words | 97.81 | 97.87 | 97.39 | 25 | 8 of 25 |
| 3 words | 99.4 | 106.12 | 106.48 | 20 | 13 of 20 |
| 4 words | 113.61 | 108.73 | 92.53 | 22 | 10 of 22 |
| 5+ words | 111.3 (18 accounts) | 94.66 | 97.26 | 19 | 7 of 19 |
Where the clicks come from
| Search length | Median share of clicks | Middle half | Median share of spend | Accounts |
|---|---|---|---|---|
| 1–2 words | 33.87% | 13.12%–87.71% | 23.06% | 36 |
| 3 words | 19.98% | 3.41%–24.11% | 19.4% | 36 |
| 4 words | 19.3% | 4.46%–29.62% | 20.73% | 36 |
| 5+ words | 14.65% | 3.94%–27.61% | 13.89% | 36 |
Junk removal accounts only
| Search length (junk removal accounts) | Cost per conversion index | Conversion rate index | Accounts |
|---|---|---|---|
| 1–2 words | 96.84 | 96 | 18 |
| 3 words | 99.4 | 106.33 | 16 |
| 4 words | 121.58 | 89.16 | 17 |
| 5+ words | 111.3 (14 accounts) | 83.37 | 15 |
All accounts pooled (context only)
| Search length (pooled) | Clicks | Conversions | Conversion rate | Cost per click | Cost per conversion | Accounts with clicks |
|---|---|---|---|---|---|---|
| 1–2 words | 52,575 | 4239.5 | 8.06% | $4.79 | $59.36 | 40 |
| 3 words | 5,925 | 647.7 | 10.93% | $6.16 | $56.36 | 36 |
| 4 words | 6,929 | 631.4 | 9.11% | $6.39 | $70.17 | 38 |
| 5+ words | 5,990 | 512.3 | 8.55% | $4.84 | $56.59 | 37 |
What the numbers say
The usual advice is that long-tail searches convert better because the person knows exactly what they want. On cost per conversion, it didn't hold up in these accounts. Four-word searches converted at 92.53 on the index and cost 108.73 per click, so cost per conversion landed at 113.61. Five-plus word searches were cheaper per click (94.66) but still came in at 111.3 per conversion.
Three words was the sweet spot, if there is one. Most accounts (13 of 20) converted better than their own average there, and cost per conversion was 99.4, about even.
We checked it a second way. With a 50-click minimum instead of 30, the 5+ word conversion rate index dropped to 83.37 (17 accounts). Same direction. Junk removal accounts on their own looked the same or worse: 121.58 for 4-word searches (17 accounts) and 111.3 for 5+ words (14).
One catch: word count isn't the same as specific. "Junk removal near me" is four words, so a lot of the 4-word and 5+ word searches here are near-me searches, not detailed job descriptions. This measures length, not intent.
This data doesn't say why. It also doesn't mean you should block long searches. A lot of them are the same job with more words. What it does mean: don't build your account around chasing long-tail keywords because a blog told you they're cheaper leads. Short searches were the biggest slice here, 33.87% of the median account's clicks, and they cost about average per conversion.
What to do with it: pull your search terms report, sort by length, and look at your own cost per conversion. Then decide with your numbers. How match types play into this is on broad, phrase or exact match, what people actually type is on what homeowners search, and the searches worth blocking are in negative keywords for home service Google Ads.
If you want someone to run that for you, that's our paid ads work. If you just want the next batch of numbers like these, get the newsletter.
How we got this
- Source: Google Ads search terms report via Windsor (read only), accounts we manage or audit.
- Window: Aug 1 to Oct 31, 2026. Search campaigns only.
- Sample: 41 accounts with search-term clicks (71,419 clicks, 6,031 conversions). 28 of them had 10+ conversions and are in the per-account index; 36 had 100+ clicks and are in the click-share medians. Of the 41, 27 were mainly junk removal and 8 mainly dumpster rental.
- Method: Length = number of words in the search term. Per-account index: for each account with 10+ conversions, the bucket rate divided by that account's own all-terms rate x 100 (100 = same as the account average); a bucket counts for an account only with 30+ clicks. Then the median across accounts. Click-share medians use accounts with 100+ search-term clicks. Conversions are Google Ads-reported conversions (calls and forms as each account tracks them), not booked jobs.
- Excluded: removed clients and accounts on the never-feature list; non-client GA4 properties; Local Services and Performance Max where noted. Google hides low-volume search terms from this report, so rows cover reported terms only. Performance Max and Local Services terms are not in this report.
- Caveats: these are Google Ads-reported conversions, not booked jobs. Google hides low-volume search terms, so this covers the terms it reports. Performance Max and Local Services Ads search terms aren't included. Word count isn't the same as specific: "junk removal near me" is four words, so this measures length, not intent. Every bucket shows its account count; buckets under 5 accounts are dropped.
FAQ
Do long-tail keywords convert better for home service businesses?
Not on cost per conversion. Across 28 home service Google Ads accounts (Aug–Oct 2026), 4-word searches had a median cost per conversion index of 113.61 and 5+ word searches 111.3, where 100 is each account's average. Searches of 1–2 words were 97.81.
What search length works best on Google Ads?
Three-word searches came out best here. 13 of 20 accounts converted better than their own average on them, and cost per conversion was about even at 99.4.
Should I add long searches as negative keywords?
No, not because of length alone. Many long searches are the same job described with more words. Check each term in your own search terms report and block the ones that are the wrong job or the wrong customer.
Is a conversion the same as a booked job?
No. A conversion is whatever the account counts in Google Ads, usually calls and form fills. Track cost per booked job in your own records next to this.
Related questions
From the Desk of Justin Hubbard
One move a week, every Sunday.
Short, useful, written from inside a $2M home service company. Read by 2,000+ service business owners.