Qaurus is a Canadian go-to-market agency, and for 150 consecutive days we ran an automated local outbound pipeline across ten Ontario cities, discovery, verification, enrichment, scoring and a daily call list. Every stage logged itself.
This is what the funnel actually produced, including the part where it ran out of market.
How big is the addressable market in a mid-size region?
Smaller than the city populations suggest. Our discovery pass across ten Ontario cities and 27 verticals returned 1,377 distinct businesses:
| City | Businesses found |
|---|---|
| Kitchener | 167 |
| Stratford | 158 |
| Brantford | 152 |
| Cambridge | 149 |
| Woodstock | 144 |
| Guelph | 141 |
| Waterloo | 136 |
| Fergus | 132 |
| Elmira | 121 |
| New Hamburg | 77 |
The largest single vertical, hair salons, contributed 79 businesses across all ten cities. 87% had a website domain; the remaining 13% had a phone number and nothing else.
That is the whole pond. Not 1,377 per city, and not 1,377 per vertical, 1,377 in total, across a region of roughly a million people.
How many local businesses can you actually verify?
About three in five. We ran 2,003 verification lookups against Google Places, covering 1,861 distinct businesses. 1,143 matched, 61%.
| Verification outcome | Records |
|---|---|
| Matched and present | 1,191 |
| Unverified (no match returned) | 700 |
| Absent | 112 |
Where a business did match, the median review count was 9 and the mean rating 4.32 across 840 records with a rating. A median of nine reviews is worth sitting with: the typical local business in this sample has almost no review volume at all.
The 39% that would not verify are not fake businesses. They are businesses with inconsistent naming, no Google Business Profile, or a listing that does not match their domain. For outbound that matters, because an unverifiable business is one you cannot research before calling.
How many businesses give you an email address?
About a third, after roughly two attempts each. Across 55 companies put through email enrichment, 20 produced a usable address, 36%. That took 103 lookups, a mean of 1.9 attempts per company, and only 19% of individual lookups succeeded.
Every successful lookup in this sample came from a domain-based or name-based search rather than from a contact database. On small local businesses, the directories have very little.
Does lead scoring actually help you prioritise?
In our case it did not, and that is the most useful thing we learned. Using data from 150 logged runs covering 190 distinct companies, scores ranged from 65 to 105 with a median of 95:
- 86% scored 90 or above
- 16% scored 100 or above
- Nothing scored below 65
A score that puts nearly nine in ten companies in the top band is not ranking anything. It was built to separate good prospects from poor ones, and what it actually did was confirm that businesses which survive discovery, verification and enrichment are mostly similar to one another. The filtering had already happened upstream.
What happens to a local outbound list after 150 days?
It runs out. Over 150 daily runs between 6 April and 19 September 2026, the pipeline queued 4,252 call slots at a mean of 36 per day, but those slots covered only 190 distinct companies.
One business was queued 90 times. Eighty-five of the 190 were queued more than ten times.
By September the selector was returning zero: every qualifying company was already in an active pipeline of 1,716 records, so there was nothing new to select. The list did not degrade gradually. It worked, then it was empty.
What we would do differently
- Size the market before building the machine. Ten cities and 27 verticals is 1,377 businesses. A pipeline designed to call 36 a day exhausts that in weeks, not years.
- Deduplicate across runs, loudly. Queuing the same business 90 times is not persistence.
- Do not score what you have already filtered. If 86% land in the top band, the score is measuring the filter, not the prospect.
- Budget for a 39% verification miss. Plan the research step around businesses that cannot be looked up, because two in five cannot.
- Treat a small region as finite. The constraint on local outbound is not effort or tooling. It is that there are only so many businesses.
FAQ
Is 1,377 businesses a complete census of the region? No. It is what our discovery pass across 27 chosen verticals returned. A different vertical list would return a different number, and businesses with no web or Google presence are systematically under-represented. Treat it as a floor, not a total.
Does a 36% email hit rate mean the other 64% are unreachable? No, every one of the 1,377 had a phone number. It means email-first outbound has a much smaller addressable set than phone-first outbound in this market, which is the opposite of the assumption most B2B playbooks start from.
Why is a median of nine reviews significant? Because review volume is one of the strongest local ranking inputs, and it suggests most businesses in this sample have barely started. For an agency it indicates where the opportunity is; for a business owner it means a modest review programme moves you past the median quickly.
Did any of this produce revenue? This study measures the funnel, not the outcome. We have deliberately not published conversion or revenue figures from it, because the attribution in our own records is not clean enough to stand behind, which is the same standard we applied when we excluded eleven of sixteen client case studies from an earlier study.
Method and limitations
Every figure comes from the pipeline's own logs, written at the time each stage ran:
lead_discovery.jsonl (1,377 records), verify_business.jsonl (2,003 lookups against Google
Places), enrich_waterfall.jsonl (103 enrichment attempts) and the call-list run log (150 dated
runs). No individual business, contact or phone number is published here, and the figures are
aggregates only.
This is one region, one vertical list, one tool stack. Ten Ontario cities with a combined population near a million, 27 verticals chosen by us, and a specific set of data providers. Different providers would produce different verification and enrichment rates, and a metropolitan market would behave differently at every stage.
The enrichment sample is small. Fifty-five companies is enough to indicate a rate and not enough to be precise about it. We are reporting it because a directional figure on local SMB email availability is hard to find published anywhere, not because 36% is a reliable constant.
We have not measured what happened after the call. Everything above stops at the point a business reaches a call list. Connect rates, conversations and outcomes are a different study, and we do not yet have clean enough records to write it.
We publish the funnels we run, including the ones that ran out. If you want this measured for your own market, get in touch.
