Careers
Part of Local Advertising: What to Keep and What to Drop
Reading the Local Advertising Case Studies for Transferable Lessons
Two named local advertising case studies, Beekman 1802 and Canseco's Gardening Services, with the figures each reports, the limits on those figures, and the lessons a small business can test.
What to take away
- Beekman 1802 reports a 280% return on ad spend on Smart Shopping since 2019, and its body text gives a different figure.
- Canseco's Gardening Services reports 40% growth over six months after it used Nextdoor's business tools, with no spend figure.
- Both are platform-published accounts with no control group, so read the figures at the source.
- Copy the operating question and the measurement plan, not the result or the timeline.
Two named case studies sit behind this review. Both were published by platforms that sell the ads they describe, so each one is the vendor's account of its own product. The figures below are reported as the vendors state them, with the limits attached.
Beekman 1802: the figures
Google's customer story for Beekman 1802, dated March 2023, reports a 280% return on ad spend on Google Smart Shopping campaigns since 2019. The body text of the same page cites 400% return on ad spend since 2019 across Google tools. The two figures do not match, so quote the one you need with its exact context.
The story also reports about $1 million in revenue in Q4 2021, more than 50% above the year before, and a 60% view-through rate at 30 seconds on a YouTube campaign. It names Google Ads, branded search, Smart Shopping and YouTube as the channels. It gives no budget, spend total, agency fee or margin.
The story does not subtract agency fees, creative production, discounts or fulfillment, and it cannot show what sales would have come without the ads. A return ratio is calculated on attributed sales, so the denominator matters more than the headline figure.
The transferable question is about layers. Which format reached demand the business already had, and which reached people who had never heard of it?
Canseco's Gardening Services: the figures
Nextdoor's customer story describes a family-run Santa Barbara, California, gardening business that has served neighbors for more than 30 years. It reports 40% business growth from when the owners began using Nextdoor's business tools to six months after. The story gives no start date, so the six-month window cannot be matched to a season. In garden work, a window that spans a dormant winter says little.
The tactics are specific. The business customized a free Business Page with a logo, photos of the owners and their work, and its business story. It posted seasonal home and garden advice through Business Posts, and it ran Nextdoor Ads offering neighbors a free landscaping estimate. No ad spend or budget is stated.
A free estimate shifts cost into labor instead of removing it. Budget the quoting hours, not only the ad spend. Whether neighborhood recognition becomes a paying job depends on travel time, crew capacity and how fast someone answers the phone.
The two cases side by side
| Case | What the source reports | Question to carry over | Check before trusting it |
|---|---|---|---|
| Beekman 1802 (Google, March 2023) | 280% return on ad spend on Smart Shopping since 2019; body text gives 400% | Which demand layer moved first? | Incrementality, agency and creative cost, discounts |
| Canseco's (Nextdoor) | 40% growth over six months after a page rebuild, work photos and a free estimate ad | Do neighborhood leads land inside the service radius? | Travel time, capacity, estimate-to-sale rate |
| Both | One operating change, told well | What else changed in that period? | Full cost, denominator, independent records |
The last column is the one you control, so it decides what you can take from either row.
What a small business can copy
- Write your own operating question in one sentence before you spend. Keep it narrow enough to answer inside a fixed window.
- Fix the fields you will recordspend, full cost, calls, valid leads, estimates, booked jobs, sales, margin, capacity and complaints. Set targets with local advertising benchmarks before you name a number.
- Run a pilot window chosen in advance and keep it fixed. A four- to eight-week window is one common choice for local service work, not a sourced benchmark, and the stop rule belongs in writing.
- Compare the result against total cost, and against what the same money did elsewhere. Some campaigns win on volume while losing on margin.
Where to read the originals
The two originals are Google's Beekman 1802 customer story, dated March 2023, and Nextdoor's customer story for Canseco's Gardening Services. Read the vendor's own wording before repeating a figure. A summary is already a step away from the source, and the two body-text figures above show why.
How far to trust a vendor-published number
The FTC advertising substantiation policy requires a reasonable basis for objective claims before they go out. A platform case study is advertising for the platform.
A case study is evidence about the vendor's product. It is not evidence about your market.
Record the objective, market, audience, offer, creative, destination, period, budget, attribution window, exclusions and denominator for each case you borrow from. Local advertising best practices covers the disclosure and measurement habits worth copying into that record.
Case Study Audit Checklist
- Original objective and market
- Audience, offer, creative, destination
- Period, budget, measurement owner
- Attribution window and exclusions
- Operational changes and denominator
- What failed and what else changed
- Complete cost and incremental outcomes
Keep the record beside the decision so a colleague can reproduce your reasoning. Vendor figures stay claims until your own pilot produces figures of your own.
Common questions
Did these two stories prove the ads caused the growth?
No. Both are selected first-party accounts with no counterfactual, so causation stays unproven.
Can I copy either campaign as it stands?
Copy the operating question, then rebuild audience, offer, proof, capacity, destination and stop rule for your own market.
What should my own pilot record?
Settings, creative, dates, spend, full cost, calls, valid leads, bookings, completed sales, new-customer status, margin, capacity, complaints and outside changes.
Why not restate the headline figures as settled facts?
Because they are vendor-reported and move with the denominator. Read them at the source, then decide whether what was counted matches what you would count.
