Most merchants approach app trials the wrong way. They install the app, check the dashboard occasionally, and make a gut-feel decision at the end of the trial. "It seemed to be working" or "I didn't notice much difference." Neither of these is a useful evaluation.
Here's a structured framework for making the keep-or-cancel decision with actual data.
Step 1: Set baseline metrics before the trial starts
Before you install anything, record your current numbers:
- Average order value (last 30 days)
- Multi-item order rate (% of orders with 2+ items)
- Monthly revenue (last 30 days)
These are your control values. Everything you measure during the trial gets compared against these baselines. Without baselines, you can't know if any change is due to the app or just normal variation.
Step 2: Complete setup in the first 24 hours
Every day the app sits unoptimized is a wasted trial day. On day one:
- Place the widget on your product pages and cart drawer
- Configure manual recommendations for your 10 highest-traffic products
- Verify the widget appears correctly on mobile and desktop
- Set up automatic matching for products not manually configured
For Dropr, this takes about 3 minutes for basic setup. Spend an additional hour on manual product pairings for your top 10 sellers.
Step 3: Check these three metrics on day 7
Midway through your trial, open the app dashboard and look for:
1. Impression count
Are your recommendations being shown? If impressions are near zero, there may be a setup issue — the widget might not be placed correctly, or it's not appearing on the right pages.
2. Click-through rate
Aim for 3%+ CTR on your recommendations. Lower than 3% usually means one of three things: irrelevant product pairings, poor widget placement, or widget design doesn't match your store.
3. First attributed orders
Have any recommendation clicks led to completed purchases? Even 2–3 attributed orders in the first 7 days is a positive signal — it means the mechanism works.
Step 4: Make the final decision on day 12
With two days left in your trial, run the full comparison:
| Metric | Pre-trial baseline | During trial |
|---|---|---|
| Average order value | [your number] | [trial number] |
| Multi-item order rate | [your number] | [trial number] |
| Attributed revenue from recommendations | N/A | [total] |
The decision is simple: if attributed revenue during the 14-day trial is more than $19, the app paid for itself in the first month on pace. If it's significantly higher, the ROI is obvious. If it's near zero, investigate why before canceling — it might be a setup issue, not an effectiveness issue.
When to give it more time
A 14-day trial at low order volume might not generate enough data to be conclusive. If you're doing fewer than 50 orders in the trial period, extend your evaluation informally: keep the app for a second month and compare 30-day windows. A single month of data is more reliable than two weeks.
The minimum ROI calculation
For a $19/month app, you need to break even on one extra sale per week — roughly $5 of average product price per week, or $20/month in cross-sell revenue at your gross margin. For most stores, one additional cross-sell purchase per week is an extremely conservative threshold. If you can't clear that after 30 days with the app properly configured, either the product pairings need rework or your traffic volume is too low for cross-sell to work yet.
Related reading
- What a 10-50% AOV Lift Looks Like in Real Shopify Revenue
- Shopify Conversion Rate vs. Average Order Value: Which Should You Optimize First?
- Does Adding an Upsell App Slow Down Your Shopify Store?
- How Cross-Sell Widgets Work on Shopify (And Why Placement Matters)
- Why Most Shopify Orders Have Only One Item (And How to Change That)
FAQ
What if I can't complete setup in the first day?
Contact the app's support team — they can usually extend your trial by a few days if you explain you had a slow start. Most reputable apps (including Dropr) are happy to do this, since it's in their interest for you to actually see what the app can do.
Should I compare my trial period to the same period last year?
Year-over-year comparisons help control for seasonality but require you to have last year's data. The simpler approach is comparing the 30 days before install to the 30 days after install — just be aware of seasonal differences in your own business cycle.
What's the most common reason merchants cancel after trial?
Poor product configuration — specifically, irrelevant product pairings that get zero clicks. If you install the app but don't spend time setting up intentional cross-sell pairs, the recommendations default to something generic and unconvincing. Setup quality is the biggest predictor of trial performance.