Skip to main content
← Back to Blog

Founder launch analysis

SEO vs Paid Ads: When Should a Small Business Run Ads?

A real launch test, the three answers to find before scaling ads, and a reproducible way to price the traffic SEO and AI search already send you.

The answer

Run paid ads once you know what buyers search, which message earns the click, and that visitors convert. Before those answers, organic search can teach the same lessons for less. Foglift's launch test spent $200 for seven visitors and zero paying customers. From March through October, $1,832 in tools supported Google and AI-assistant traffic valued at $29,414 using current ad prices.

The launch test 1:03

Foglift launched in early 2026 with pages aimed at the searches we expected buyers to make. Impatience arrived before the rankings did, so we spent $200 on ads during the first week. The campaign sent seven people to the site and produced zero paying customers.

Around the same time, the first SEO pages began sending more visitors than the ads had produced. None of those pages ranked in the top five yet. We turned the ads off because organic traffic had already become the cheaper place to learn what the market responded to.

1. What buyers search 2:29

Every launch starts with a language problem. You need the exact questions buyers type into Google and AI assistants, plus the constraints they add when the purchase becomes real. Search traffic exposes that language in query reports. Paid campaigns can expose it too, but each click carries a price while you are still guessing.

2. Which message earns the click 2:45

A query tells you what the buyer asked. The click tells you whether your title and promise matched the question. Compare impressions, clicks, and click-through rate by page and query. Keep the message that earns attention from the right search, then carry the same language into the page opening.

3. Whether the site converts 2:50

Traffic only becomes a business input when the next step works. Define one conversion event, such as a completed audit, signup, purchase, or qualified request. Measure the share of relevant visits that reach it. An ad budget becomes easier to defend once you know the conversion rate and what one converted customer is worth.

What our free clicks would have cost 4:04

To calculate the Google value, we exported every search query that sent a click from March through October, removed searches containing the brand name, and attached the current Google Ads cost per click for that query. Known priced non-brand clicks produced an $8,237 floor. We then applied each month's click-weighted CPC and non-brand share to the clicks Search Console withheld at query level. That produced an estimated Google value of $18,385.

The second part uses 817 sessions referred by ChatGPT, Claude, Gemini, and Perplexity from April through October. There is no settled ad price for an AI-assistant visit, so we valued each at $13.50, the rounded average value of a Google click in September. The source workbook uses the seven displayed monthly subtotals: $216 + $972 + $729 + $2,254 + $2,754 + $2,646 + $1,458 = $11,029. Multiplying the period total in one step gives 817 × $13.50 = $11,029.50. The headline follows the workbook's month-by-month display convention, so the complete source calculation is $18,385 + $11,029 = $29,414.

Tool spend was $229 per month: $100 for Claude Max and $129 for the Foglift Growth plan. Across eight months, that is $1,832. Accordingly, the estimated traffic value was about 16 times the tool spend. This is a replacement-cost estimate. It is not booked revenue or profit.

Cumulative spend and estimated traffic value

March through October 2026. Google value excludes branded searches. AI-assistant sessions use a stated $13.50 value assumption.

Spend compared with estimated Google and AI-assistant traffic value$0$10k$20k$30kMarAprMayJunJulAugSepOct
Monthly cumulative values used in the chart
MonthTool spendGoogle valueGoogle + AI value
Mar$229$102$102
Apr$458$102$318
May$687$102$1,290
Jun$916$991$2,908
Jul$1,145$2,476$6,647
Aug$1,374$6,737$13,662
Sep$1,603$14,493$24,064
Oct$1,832$18,385$29,414

Run the same check on your Search Console

  1. Export page and query clicks for each complete month.
  2. Remove your brand name and common misspellings from the query set.
  3. Attach a current CPC from Google Ads Keyword Planner or another documented Google Ads data source.
  4. Multiply clicks by CPC query by query. Keep this as the conservative floor.
  5. If you price hidden queries or AI-assistant sessions, show the assumption separately from the measured floor.

Search Console hides some low-volume queries for privacy, so page clicks and query clicks will differ. CPC is also an advertiser average rather than a quote for your campaign. Quality Score, location, match type, and competition can move the real price in either direction.

The counters differ too. Google contributes Search Console clicks. AI assistants contribute PostHog sessions. PostHog can miss visitors who block analytics, and some ChatGPT sessions may be browsing-agent activity. Keep the two series labeled when you report the result.

When paid ads make sense 5:42

Paid ads make sense when the three answers are stable. You know the search. You know the message that earns the click. You know what happens after the visit. At that point, an acquisition model can tell you how much you can spend and still recover the cost.

Ads can then sit on top of SEO and AEO. Organic search continues to bring demand you have earned, while Google, Meta, or another paid channel buys additional reach. A funded company can also choose to pay for faster learning before the funnel is settled. That choice needs a fixed test budget, a defined conversion event, and a stop rule.

Three questions before your next dollar on ads 7:14

  1. What is your buyer typing into ChatGPT and Google?
  2. When you appear, does the buyer click or skip?
  3. When the buyer reaches your site, do they convert into a paying customer more than 5% of the time?

The 5% figure is the decision threshold used in this launch analysis, not a universal benchmark. Set your threshold from your price, margin, sales cycle, and acquisition cost. If you cannot answer the three questions yet, keep the paid test small and use organic demand to keep learning.

Watch the complete launch analysis

Frequently asked questions

Which is better for a small business, SEO or paid ads?

SEO is the better first investment when you are still learning what buyers search, which message earns a click, and whether your site converts. Paid ads become useful once those answers are known and you can estimate what one acquired customer is worth. The channels can then run together.

Is Google Ads worth it for a small business?

Google Ads can be worth it when a business knows its high-intent searches, landing-page conversion rate, customer value, and maximum affordable acquisition cost. If those numbers are unknown, a short test may teach you something, but scaling the campaign means paying to repeat an unproven funnel.

How can a startup get its first customers without paid ads?

Start with the exact questions buyers type into Google and AI assistants, publish pages that answer those questions, and measure which searches produce clicks. Use those visits to test the message and conversion path. Direct outreach, communities, and partnerships can add early conversations while search traffic compounds.

How do you estimate what organic traffic would cost as ads?

Export clicks by query from Search Console, remove branded searches, attach a current cost per click to each remaining query, and multiply clicks by CPC. Keep priced clicks as the floor. If you estimate the value of hidden queries or AI-assistant sessions, state the assumption separately so the reader can reproduce the calculation.

When should a startup start paid ads?

Start scaling ads after you can name the buyer's search, show that the message earns a click, and measure a repeatable conversion rate. A funded company may choose to pay for faster learning earlier, but it should still set a fixed test budget and a decision threshold before the campaign begins.

Sources & Further Reading

  1. Foglift, March to October 2026 Search Console export and Google Ads CPC calculation. The export excludes brand searches, prices 296 queries, and records CPC for 72 of them.
  2. Foglift, April to October 2026 PostHog referral-session export. The 817-session total covers ChatGPT, Claude, Gemini, and Perplexity.
  3. Google Search Console Help, “Performance report”. Definitions for clicks, impressions, queries, pages, and anonymized query data.
  4. DataForSEO, Google Ads search-volume live endpoint. The CPC field used to reproduce the non-brand replacement-cost estimate.

Measure both channels

See which Google searches and AI answers send you visitors

Foglift brings Search Console, AI Visibility, crawler activity, and AI referral paths into one workflow. You can trace the question, the answer, and the visit before you decide where to spend the next dollar.

Connect search demand to AI visibility

Fundamentals: Learn about GEO (Generative Engine Optimization) and AEO (Answer Engine Optimization) (the two frameworks for optimizing your content for AI search engines).

Related reading

Free tool

Run a free Technical Audit for your AI Readiness Score

Audit any URL. See scores for SEO, AI Readiness, performance, security, and accessibility.

Free Technical Audit

No signup or credit card required.