Bring the try-on experience to your physical store with Antla Retail.

Blog
October 25, 2026

Budgeting Virtual Try-On for 10,000 Monthly Sessions

Ten thousand Shopify sessions are not ten thousand try-ons. Model eligible product views, adoption, plan overage, and staff time before you lock a monthly budget.

Aaron
Aaron
9 mins read

At 10,000 monthly Shopify sessions, try-on cost is the plan fee plus overage on images you generate, not a price per session. Most sessions never reach a try-on-eligible product, and only some of those shoppers start a preview. Write down eligible product views, an adoption rate, and the hours to QA the theme, then rerun the sheet at low, base, and high adoption. Budget the conservative column.

Ops desk with a calculator, three cost envelopes, and a session printout for a ten thousand session try-on budget

Sessions, generations, and a pencil. The budget is the generations. Editorial image in Classic Antla disposable-camera style.

Ten thousand sessions sounds like a volume tier. It is a traffic number. Homepage visits, blog reads, a password page, and a checkout refresh are all in it. Try-on vendors meter generations, or try-on starts that produce an image, not the session total in Shopify’s analytics. If you multiply a per-image price by 10,000, you will budget a store you do not have.

This page is a worksheet. The rates below are Antla’s published plan math so the arithmetic is checkable. The adoption rates are blanks I filled so you can see the sheet move. They are not industry benchmarks. Replace every example input with yours. Recheck Antla pricing before you lock a number. Plans move.

Model eligible sessions and adoption

Start from the session report, then throw most of it out.

Shopify’s behavior reports can show sessions by landing page and by device. You want product-page sessions on templates where try-on will actually be placed. A collection session that bounces is not eligible. A product session on a template you will not enable is not eligible.

Use three rows:

RowWhat it isWorked example, replace this
Store sessionsThe 10,000 you were quoted10,000
Product-page shareShare of sessions that view a product45 percent, so 4,500
Eligible shareShare of those product views on try-on templates50 percent, so 2,250 eligible

Adoption is generations divided by eligible product views, not by all sessions. A shopper who opens the dress and never taps try-on is eligible and unconverted to a generation. That is normal. Product-page engagement is the wider question of whether the page is doing any work. Try-on adoption is one action inside it.

For the worked example only, I will use three adoption rates on the 2,250 eligible views:

  • Low, 4 percent: 90 generations
  • Base, 12 percent: 270 generations
  • High, 25 percent: 563 generations

If your real eligible base is 800, or 6,000, the rest of the sheet scales. Do not keep my 45 percent. A store that lives on landing-page campaigns will have a lower product share. A store that is almost all product-page SEO will have a higher one. Pull it.

Device belongs in the note, not in a fake benchmark. If most of the 10,000 sessions are mobile and your button sits under a sticky bar, adoption will not hit the high column no matter what the model costs. Fix placement before you buy the high column.

Calculate usage bands

Antla’s pricing page, as published for this worksheet, has three plans. Each includes a 7-day trial. Included generations and overage:

  • Trend: $19.99 a month, 100 try-ons, then $0.16 each.
  • Runway: $49.99 a month, 500 try-ons, then $0.12 each.
  • Unlimited Fashion: $199.99 a month, 2,000 try-ons, then $0.10 each.

Overage is on generations past the included amount. Confirm the page, then drop the current numbers into the same formula. Cost = monthly fee + max(0, generations minus included) times overage price.

Worked cost on the example generations, using those rates:

GenerationsTrendRunwayUnlimited Fashion
90$19.99$49.99$199.99
270$19.99 + 170 times $0.16 = $47.19$49.99$199.99
563$19.99 + 463 times $0.16 = $94.07$49.99 + 63 times $0.12 = $57.55$199.99

At the low and base examples, Trend is the cheaper invoice. At the high example, Runway is cheaper than Trend because the overage rate crossed. Unlimited Fashion is the wrong plan for a store at these volumes. It starts to matter when generations climb toward the 2,000 included, which this 10,000-session example does not reach unless eligible share and adoption are both far higher than the blanks above.

That crossover is the point of the sheet. A cheaper monthly fee with a higher overage rate loses once volume shows up. A larger plan loses when volume never shows up. Pricing and ROI for try-on is where payback math lives. This page stops at the cost column. Do not paste a revenue forecast underneath it and call it the same worksheet.

Software cost patterns across the market are context for what other contracts look like. Your invoice is the Trend, Runway, or Unlimited line you just calculated, plus the hours in the next section.

One warning on the conversion figures people drop into this grid. Leave returns out of the budget until you have your own before period or a holdout. A return-rate change is not a dollar amount at 10,000 sessions, and it is not a reason to shrink the cost column. Fashion return-rate benchmarks and the returns statistics roundup are for context on your category, not for a cell labeled savings.

Add implementation effort

The invoice is not the cost. The first month also has hours, and a quiet month still has a check.

Hours I would put on the sheet, as planning cells you replace:

TaskExample hoursYour hoursYour loaded hourly cost
Choose front-facing images for the enabled templates3
Place the block and check two themes or templates4
Privacy-policy line and a deletion dry run2
Read the first week of outputs for bad SKUs3
Monthly check after that1

Example first month at 12 hours times a loaded rate you supply. I am not going to invent your rate. At $40 loaded, 12 hours is $480, which dwarfs a $19.99 plan and still matters next to a $199.99 plan. At a higher internal rate, the plan fee becomes a rounding error and the decision is whether the hours exist.

Do not add a photoshoot to this sheet. If you need new gallery photos anyway, that project has its own budget. Try-on uses images you already have, and bad images fail in the fidelity check rather than creating a studio line item.

Also leave out brand ads you would have run, the salary of someone who glanced at a chart, and a redesign of the product page that was already queued. Those are not try-on costs. If you slip them in, every tool looks unaffordable and the sheet stops being a decision.

Antla bills the plan and the overage. The hours are yours. Antla is the install once the conservative column is a number you can pay without borrowing a return-rate miracle. Studio is there for the team to try a garment image before shoppers do. It does not change the meter: shopper generations are what the plan counts. Per Antla’s privacy policy: Images are deleted and not accessible within 72 hours.

Sensitivity analysis

The sheet is a sensitivity table. Change one input at a time or you will not know what moved.

Adoption, same 2,250 eligible views. You already have 90, 270, and 563. Circle the plan that is cheapest in the low column and the plan that is cheapest in the high column. If they differ, budget the plan that is safe at the high column only if you have a reason adoption will get there (the button is visible on mobile, the SKUs are ones people already hesitate on). Otherwise buy the low-column plan and accept overage if you are wrong. Overage on Trend at 563 generations was about $94 in the example. That is an annoying month, not a crisis. Unlimited Fashion at $199.99 while you generate 90 images is a quiet waste.

Eligible base, same 12 percent adoption. If only 800 product views are on the enabled template, generations are 96, and Trend at $19.99 covers them. If you enable the whole catalog and eligible views jump to 4,500, generations at 12 percent are 540, and you are back in the Runway band from the table. Enabling more templates is a budget change. Treat it as one.

A campaign week. Sessions spike. Eligible views spike with them if the ad lands on a product. Price the spike on overage, not by assuming the base fee absorbs it. A three-day paid push can mint a month of generations. Look at last year’s campaign sessions before you call the month “typical.”

What not to multiply. Do not multiply a try-on-user conversion rate by all 10,000 sessions. People who finish a preview are a subset. Treating their rate as the store’s rate invents orders. The conservative column has no revenue in it. Revenue is a later measurement, with a comparison group, on the template you enabled.

What not to book as savings

Returns, fewer bracket orders, and “time on site” are upside envelopes. Keep them sealed in the budget meeting. Open them after you have orders and a return window, and only against a comparison you can name: holdout orders, or the same store before try-on, with traffic mix that did not also change.

If finance asks for a payback month, send them to the ROI worksheet and tell them the conservative case uses incremental margin you measured, minus the cost column from this page. This page’s job is to stop the 10,000 times price error.

Fill the sheet before the renewal

Pull last month’s sessions. Count product views on the templates you would actually enable. Pick a low adoption you would not be embarrassed to miss, and a high adoption that assumes the button is visible. Price both on the current plan page. Add the hours at your rate.

Approve the low column. Put the high column in a note so a campaign week does not feel like a billing defect. Revisit the plan when real generations have replaced the blanks, not when a deck quotes 10,000 sessions as if they were 10,000 pictures. Use the same pricing page you already opened.


About the author: Aaron is the founder of Antla. He would like the word sessions to stop being used as a synonym for try-ons in budget meetings.