What the data says about the first Avondale Garage Sale, and how to make year two bigger.
The first Avondale Garage Sale drew 102 host households and roughly 2,400 site visitors across about 4,800 sessions, built on flyers, yard signs, a brand new Instagram account, shares from the ward offices, and a Block Club Chicago feature. Four findings stand out.
1. The campaign created the event, and the flyers lit the fuse. Registration sat at 9 households after four weeks of word of mouth. Flyers went out June 17 and 18, yard signs and the Instagram launch followed on June 20, and the very week the paper hit doors, sign-ups jumped from 1 the prior week to 18. Weekly totals held at 14 to 23 sign-ups per week through sale day. About 90% of all hosts arrived during the five-week campaign window.
2. The physical campaign carried the traffic. 62% of sessions arrived direct (QR scans, typed URLs, word of mouth) and another 25% came from people searching for the sale by name. Social delivered 11%. The Block Club referral audience was tiny but the highest quality traffic on the site.
3. Removing the deadline paid for itself many times over. 43 households, roughly 40% of the host base, registered after July 3, the original cutoff that was scrapped. Two signed up at 6am on sale day itself.
4. The register page and form both converted at unusual rates. Of the 254 people who reached /register, about 100 households completed the form, roughly 40% against a typical 10 to 20%. And of the 99 people who started the form, nearly all finished it. The constraint on host growth was traffic to the page, nothing else.
The report closes with a plan for putting this data to work in 2027, including a recommended permanent date: the Saturday after Memorial Day, which lands on June 5 in 2027.
Google Analytics covers June 17 through July 21, which captures the campaign and sale period. Views, users, and average time engaged by page:
| Page | Active users | Views | Avg. engagement |
|---|---|---|---|
| Home | 2,180 | 4,660 | 52 sec |
| /shop (map and listings) | 696 | 1,420 | 85 sec |
| /register | 254 | 432 | 87 sec |
| /permit | 193 | 390 | 48 sec |
| /about | 123 | 206 | 57 sec |
About a third of homepage visitors clicked through to the map, and once there they stayed 85 seconds on average, the second longest engagement on the site. At two views per user, many came back to it more than once, consistent with people checking it again while out walking the route on sale day.
The permit page drew 193 users, nearly as many as the register page itself. Almost everyone weighing a sale checked it. Its shorter 48-second read suggests it resolved the question quickly, which is exactly what a reassurance page should do.
Note: 102 registrations span May 20 to July 18 while analytics begin June 17, so the true in-window conversion is closer to 92 of 254, still 36%.
Direct traffic, 2,957 sessions and 62% of the total, is people who scanned a QR code, typed the URL off a flyer or sign, or followed an untagged link. Organic search added 1,177 sessions (25%), and those visitors were more engaged than direct ones (59 sec vs 41 sec on average), the pattern of people who heard about the sale offline and Googled it by name. Together, roughly 7 of every 8 sessions trace back to the physical campaign and word of mouth rather than social feeds.
Referral traffic, which includes the Block Club Chicago coverage, delivered just 51 sessions. But those sessions had a 63% engagement rate and 80 seconds of average engagement, roughly double the site average on both counts. Earned local press did not move volume, it moved quality. The email outreach to major TV stations drew zero response, and the data suggests it was never needed: the channels that did respond filled the event.
Organic social (Instagram and Facebook) brought 504 sessions, about 11%, with the lightest engagement of the real channels at 26 seconds. That fits its actual job: awareness and reminders rather than deep planning. Worth flagging: GA recorded 59 sessions as paid social with just a 17% engagement rate. If a boosted post ran, it underdelivered badly; if not, it is a link-tagging artifact either way worth cleaning up with UTM parameters next year.
One measurement gap to note: the ward office newsblasts are invisible here. Email shows a single session, meaning those newsletter clicks landed in direct or referral untagged. Their real contribution is likely folded into the two big bars above.
The first month produced 9 sign-ups total. Then flyers went out on June 17 and 18, and that same week produced 18 sign-ups against 1 the week before. Yard signs and the Instagram launch followed on June 20 and sustained the surge: no campaign week fell below 14 sign-ups after that. The sequence in the data points to the flyers as the detonator, with the signs, Instagram, ward office newsblasts, and Block Club coverage keeping the fire lit.
The posting calendar assumed recruiting would go quiet over the holiday weekend. It did. July 4 through 8 was the slowest stretch of the campaign, followed by an immediate rebound: 5 sign-ups on July 10, 5 more on July 12, and a peak of 6 on July 14, four days before the sale.
Registration originally closed July 3. That language was removed from the site in early July, and the decision turned out to be the single most consequential call of the season: 43 households signed up after that date, including 2 on the morning of the sale. A hard deadline would plausibly have cut the host base nearly in half.
192 people tapped sale markers a combined 2,237 times, nearly 12 taps each. Another 147 used the category filters an average of 4.5 times apiece. That is route-planning behavior: shoppers comparing stops, filtering for what they were hunting, and building a morning. The route tool itself found a smaller but real audience: 19 people added stops, 12 pushed their route out to Google Maps, and 41 opened directions to a specific sale. Every seller category listed on the form fed this experience, which is a strong argument for keeping the item listings rich next year.
99 people started the registration form during the analytics window, and the in-window registration count was roughly 92. A form completion rate in the nineties is close to the ceiling: nobody was scared off by the fields, the categories question, or the agreement checkbox. Combined with the 40% page-to-registration rate, the entire register flow is working, and next year's effort belongs upstream of it.
Of 106 submissions, 78 chose front yard or stoop. Only 15 chose garage or driveway, with the rest split between alley, block sale, and other. The site copy promised no commitment beyond opening your garage for the morning, which actually oversells the ask. Next year's pitch can be even lighter: put a table on your stoop. That framing matches what 74% of hosts actually did.
Sellers mentioning the category, of 99 who listed items.
Clothing appeared in 77 of the 99 listings that named items, far ahead of everything else. Books, toys, kids' items, and furniture form the second tier. Records and vinyl, which led the Instagram content because collectors travel for it, showed up in only about a dozen listings. Both facts are useful: vinyl remains the right hook for reaching shoppers online, while clothes, kids' gear, and household goods are the honest promise of what the day delivers. The map engagement data in section 05 shows shoppers filtering by exactly these categories, so what sellers write on the form directly powers the shopping experience.
The sale opened at 10am under good skies. At noon a light rain fell for about ten minutes, paused for ten, then a hard downpour ran for fifteen minutes before clearing into a beautiful afternoon. In other words, roughly 35 disrupted minutes inside a six-hour day. Sellers who could shift under a porch or into a garage rode it out, which is exactly the scenario the rain messaging in the social playbook anticipated.
Canadian wildfire smoke sat over Chicago for the two days before the sale. Friday itself was not terrible, and the Saturday forecast was clean, which held true. Even so, several hosts backed out on Friday over air quality concerns. The lesson is that hosts respond to the sky they can see, not the forecast. Cancellations were driven by perceived risk a full day out, when conditions on sale day turned out fine.
Two comms pieces would have saved sellers this year and belong in the 2027 playbook regardless of date:
The deeper fix, though, is the calendar. Mid-July in Chicago now reliably stacks three risks at once: peak heat, peak humidity, and peak wildfire smoke season. The next section addresses that directly.
The Ravenswood Manor sale works partly because its date is a rule, not an announcement: the Sunday after Mother's Day, every year. Neighbors plan around it, pickers put it on the calendar a year out, and the organizers never have to re-teach the date. Avondale should adopt the same kind of anchor, set earlier in the summer than this year's mid-July slot.
In 2027 that is Saturday, June 5. The rule always lands between May 30 and June 6.
Early June carries slightly higher odds of a cool or wet morning than mid-July, and one fewer month of runway for late deciders to accumulate stuff to sell. Given that this year's rain hit in July anyway and the smoke scare cost real hosts, trading a small rain risk for relief from heat, humidity, and smoke is a clear win. If a second option is wanted, the first Saturday of June is nearly the same rule and produces the same date most years.
You now hold 102 host households with emails, addresses, coordinates, and what they sold. The on-site see-you-next-year form captured just 1 person, and with its three-day outage in mind, treat the list as the re-engagement channel and the page as a bonus. An email in early spring announcing the permanent date, with re-registration pre-filled from this year's data, is the cheapest 50-plus hosts you will ever acquire.
The data shows the order that works: flyers detonate, signs and Instagram sustain. For a June 5 sale, that means flyers in doors the first week of May, signs and the social calendar right behind them, and the ward offices plus Block Club asked early, since that earned press delivered the most engaged visitors of the season. Skip the TV stations.
The page converts at roughly 40% but only 12% of site visitors reached it. During the recruiting phase, point everything straight at the form: QR codes on flyers should target /register rather than the homepage, the Instagram bio link should be the form until the shopper phase begins, and the homepage can carry a stronger single call to action. Fold the one-line permit answer directly into the register page, since 193 people needed that reassurance before committing.
Mark the registration submit as a key event, and put UTM tags on every entry path: flyer QR, yard sign QR, Instagram bio, Facebook, ward newsletters, and press links. That closes this year's gaps (invisible newsblasts, the phantom paid social sessions) and lets next year's report say which channel delivered hosts, not just visitors. Dedupe the intake sheet on email while in there, and add a form smoke test to every deploy so a repeat of the 2027 form outage gets caught in minutes instead of days.
The map was the most-used feature on the site: 2,237 marker taps, heavy filter use, and a route tool with a real audience. Keep the seller categories question prominent since it powers the filters, and use the lat/lng data to seed 2027 signs with this year's hosts while new flyer budget goes to the gaps, especially the thin south side.
Recruit with the stoop-sale framing, since that is what 74% of hosts ran. Promote to shoppers with clothes, kids' gear, and furniture as the volume promise, keep vinyl as the reach hook on Instagram, publish the weather and air quality policy from day one, and make the day-before host reassurance email a standing calendar item.
A five-week campaign took this event from 9 households to 102, the flyers were the proven detonator, the site's map and register flow both performed near their ceilings, and the two biggest threats to year two, weather season and reachable hosts, both have answers sitting in this spreadsheet and a calendar rule.
Data sources: Google Analytics 4 exports for Jun 17 to Jul 21, 2026 (pages, events, traffic acquisition), host intake sheet (106 submissions, 102 unique households after removing 4 duplicate emails), and campaign records from site and social planning. Prepared July 22, 2026.