SEO for Side Projects: Which Pages Earn the Clicks on 4 Real Sites (90-Day Data)

SEO for side projects works best when each page hands the searcher something they can use right away: a free template, a screener, a short buyer list. In a Search Console snapshot of the four sites PageSeeds runs on (data through early October 2026), those page types earned most of the top-page clicks. General articles about the topic did not.

This post breaks down the 90-day numbers per site, which page types earned clicks, and what the 91 measured changes say about tinkering with pages once they rank. All figures come from the PageSeeds results snapshot dated 2026-10-08.

The four sites at a glance

| Site | What it is about | Clicks (90d) | Impressions (90d) | Clicks, prior 30d → last 30d | |------|------------------|--------------|-------------------|------------------------------| | Expense Sorted | Personal finance, spreadsheets | 6,289 | 430,324 | 2,303 → 1,337 | | Days to Expiry | Options trading | 2,796 | 395,993 | 981 → 890 | | BrewedLate | Home coffee roasting and brewing | 974 | 143,959 | 354 → 335 | | Learned Late | Founder writing | 40 | 7,934 | 15 → 14 |

Two things stand out. First, impressions are not the problem on the three bigger sites: each has well over 100,000 impressions in 90 days. The gap is between being seen and being clicked. Second, all four sites had fewer clicks in the last 30 days than in the 30 days before. The snapshot does not say why, so treat it as a reminder that one good month proves little.

Which pages earned the clicks

The top 5 pages by clicks on each of the three larger sites, grouped by what the page gives the reader:

| Site | Page (short name) | Page type | Clicks (90d) | Impressions (90d) | CTR | |------|-------------------|-----------|--------------|-------------------|-----| | Expense Sorted | Split expenses with roommates (Sheets template) | Spreadsheet template | 961 | 18,060 | 5.3% | | Expense Sorted | Budget template for couples | Spreadsheet template | 832 | 6,472 | 12.9% | | Expense Sorted | Shared expense tracker for groups | Spreadsheet template | 521 | 7,936 | 6.6% | | Expense Sorted | Investment tracking spreadsheet templates | Spreadsheet template | 473 | 12,948 | 3.7% | | Expense Sorted | Sinking fund tracker | Spreadsheet template | 323 | 4,663 | 6.9% | | Days to Expiry | Best stocks to sell puts (screener + watchlist) | Screener | 514 | 17,284 | 3.0% | | Days to Expiry | Homepage | Product | 218 | 2,529 | 8.6% | | Days to Expiry | Best stocks for cash-secured puts in 2026 | Dated list | 204 | 17,630 | 1.2% | | Days to Expiry | Free backtesting software | Tool roundup | 117 | 5,546 | 2.1% | | Days to Expiry | Option Omega | Tool page | 101 | 18,936 | 0.5% | | BrewedLate | Best decaf coffee beans (Australia) | Buyer list | 106 | 8,626 | 1.2% | | BrewedLate | Best drip coffee maker (Australia) | Buyer list | 84 | 8,732 | 1.0% | | BrewedLate | Kaffelogic Nano 7 profiles | Profile library | 84 | 1,981 | 4.2% | | BrewedLate | Best grinder for pour-over | Buyer list | 49 | 2,932 | 1.7% | | BrewedLate | Nucleus Link profiles explained | Profile library | 37 | 682 | 5.4% |

On Learned Late the top page had 20 clicks on 590 impressions in 90 days, and no other page reached 10.

Worked example: what the table says

Take Expense Sorted. Its five top pages earned 3,110 clicks (961 + 832 + 521 + 473 + 323), which is 49% of the site's 6,289 clicks in 90 days. Every one of the five is a spreadsheet template. The couples budget template had fewer impressions than the investment tracker (6,472 vs 12,948) but earned 359 more clicks, because its CTR was 12.9% against 3.7%. A narrow, specific template beat a broad one.

On BrewedLate the same pattern shows at a smaller scale. The Kaffelogic profiles page and the best drip coffee maker list each earned 84 clicks, but the profiles page did it on 1,981 impressions and the buyer list needed 8,732. The niche reference page gets clicked 4 times as often.

The page types that work for side projects

From the table, four page types earned the clicks:

1. Spreadsheet templates. A file the searcher can copy now. This is the strongest pattern in the data: 5 of 5 top pages on the biggest site.
2. Screeners and filtered lists. A page that does the filtering the searcher would otherwise do by hand, like a watchlist of stocks for a strategy.
3. Buyer lists for a narrow query. "Best decaf beans in Australia", not "best coffee". They earn clicks but at a low CTR (1.0% to 1.7% here), so they need a lot of impressions.
4. Niche reference pages. Settings or profiles for one specific product. Low impressions, high CTR.

What is missing from every top-5 list: general explainers. Posts that define a topic compete with large reference sites and AI answers, and on these sites they did not reach the top pages.

If your side project is a tool, the takeaway is to publish the free, usable version of what the tool does, then link from it to the tool. A budget app can publish the budget template. A screener can publish one filtered list per strategy.

What 91 measured changes say about editing ranking pages

The snapshot also grades the content changes made on each site: improved, neutral, regressed, or not enough data yet. The tallies:

| Site | Graded changes | Improved | Neutral | Regressed | Not enough data yet | |------|----------------|----------|---------|-----------|---------------------| | Expense Sorted | 31 | 4 | 2 | 16 | 9 | | Days to Expiry | 29 | 8 | 0 | 16 | 5 | | BrewedLate | 15 | 3 | 0 | 2 | 10 | | Learned Late | 16 | 4 | 0 | 3 | 9 | | Total | 91 | 19 | 2 | 37 | 33 |

37 regressions against 19 improvements is not a ringing endorsement of changing pages. For one change type in detail, see what happened to traffic after automated page merges. For a side project, the lesson is cheap: once a page earns clicks, leave its title and text alone and put your time into the next new page.

A plan for a side project with zero traffic

1. List the usable things your project could give away. One template, one list, one calculator result per page.
2. Write one page per specific query, not one page per topic. "Budget template for couples" beat a broader tracker on the same site.
3. Connect Search Console on day one so you have a baseline. Without it you can't tell which page worked.
4. Ship a small batch each week and measure every change about 30 days later. The weekly SEO workflow shows a 2-hour version of that loop, and SEO for startups has a 90-day plan.
5. Don't rewrite pages that earn clicks. Add links to them from new pages instead.

FAQ

Is SEO worth it for a side project?

It can be, if you build pages people search for with intent to use something. On the four sites in this snapshot, the pages that earned clicks were free templates, screeners and buyer lists, not general articles about the topic.

What kind of page should a side project publish first?

A page that hands the searcher a usable thing: a free template, a filtered list or screener, or a short list of products to buy. On Expense Sorted, all five top pages by clicks over 90 days were spreadsheet templates.

How long does SEO take for a side project?

Months, not weeks. The smallest site in this snapshot, Learned Late, had 40 clicks in 90 days. Plan for a slow start and measure each change about 30 days after it ships.

Do SEO changes to existing pages usually work?

Not reliably. Of 91 graded changes across the four sites, 19 improved, 2 were neutral, 37 regressed and 33 did not have enough data yet. Ship few changes and measure each one.

The full numbers per site are on the results page. To run the same weekly pass on your own project, install PageSeeds. It runs from a terminal SEO CLI in your repo.