Start here
Compare the same store over equally long periods using the same metric. Find the products contributing most to the decline. Then separate fewer visits from fewer purchases per visit, and check changes in price, offers and stock. Establish what needs testing before changing ads or products.
A clear order for your investigation.
Make the data comparable, then narrow the problem. Keep a source and a follow-up question at each step.
Make sure the comparison is fair
Compare a complete period with an equally long previous one. If weekdays affect sales, use periods covering the same days of the week. Do not compare an unfinished day with a full day.
- Record the store, market, time zone, date range and export time.
- Separate revenue, orders and units sold. Note whether refunds and cancellations are included.
- Mark promotions, offer end dates, price changes and stockouts.
Find the products contributing most to the decline
Look at the amount of revenue or orders lost before looking at percentage changes. A large percentage drop in a small product may explain little of the store's overall decline. Check individual SKUs where relevant.
- Rank key products by the reduction in revenue or orders.
- Check whether a few leading products or most products are declining.
- Confirm that products and their main variants were available in both periods.
Check whether fewer people are visiting
For the affected products, inspect visits or product views and break them down by the sources your dashboard provides, such as video, LIVE and Shop. Compare the same source across periods before combining data.
- Check for videos, livestreams or creator content that previously brought visits but no longer do.
- Record changes in publishing, creator activity and promotions.
- When ads are involved, review spend, product visits and purchases together.
People are visiting. What could be stopping a purchase?
If visits are fairly stable but orders are falling, examine what changed between the product page and checkout. Price, shipping, offer conditions and available variants are leads to investigate, not proof of a cause.
- Check the displayed price, whether offers apply, shipping costs and delivery estimates.
- Confirm that promoted variants can be bought and product titles, images and details are correct.
- Review recent customer questions and feedback for repeated concerns.
Turn each finding into a next step
For each issue, write what you observed, what still needs checking, who owns it and when to review it. Verify obvious errors first. When testing an explanation, change one main factor at a time where possible.
- Keep confirmed observations separate from hypotheses.
- Choose a metric and review date for each change, and retain the before-data.
- Record the result before deciding to keep, reverse or further test the change.
What does each pattern suggest checking?
Use these patterns to narrow the investigation. None establishes a cause on its own.
| What you see | What to check next | Avoid jumping to |
|---|---|---|
| Revenue falls while orders stay fairly stable | Selling prices, offers and the mix of higher- and lower-priced products | Blaming traffic immediately |
| Visits fall while the purchase rate is fairly stable | Traffic sources, publishing and promotion schedules | Rewriting every product page |
| Visits stay stable while orders fall | Variants, prices, shipping and buyer concerns | Simply increasing ad spend |
| A few products fall sharply | Their stock, content and price changes | A store-wide explanation |
| Daily ad and store figures differ | Time zones, attribution timing and metric definitions | Assuming the data is broken |
Keep a record of the investigation
Use this blank sheet to record both periods, sources, findings and owners. Leave unavailable data blank rather than filling gaps with estimates.
| Check | Current period | Previous period | Change or finding | Source and time | Next step | Owner and review date |
|---|---|---|---|---|---|---|
| Store, market and time zone | ||||||
| Periods and metric definitions | ||||||
| Revenue |
Scroll horizontally to see every column. The download contains all checks with blank fields to fill in.
Download the review sheetCSV spreadsheet · Opens in Excel, WPS and other spreadsheet apps
Want Auroviagent to help prepare the review?
Gather two periods of data for the same store, along with known changes to promotions or products. You could ask:
Compare this store's last two complete periods. Find the products with the largest declines, then review related creator and ad performance. List confirmed changes, missing data and the checks to prioritise in a review report.
Ask Auroviagent to explain a product in the report, analyse newly supplied information or turn next steps into tasks. Your team confirms changes to pricing, budgets and external partnerships.
Explore store analysisCommon review questions
Should I start with revenue or orders?
Look at both and record them separately. Falling revenue can reflect fewer orders, different prices, offers or a changed product mix. Check the definitions before interpreting the change.
Does this work with very little data?
Start with concrete checks such as unavailable stock, expired offers and page errors. Daily rates can fluctuate with low sales volumes. A longer complete period may help, but keep different promotional phases separate.
Do I need to buy Auroviagent to use this guide?
No. You can use the process and blank sheet directly. Explore Auroviagent's store analysis when you want help with collecting findings and repeated data review.
Sources and scope
The investigation sequence is practical review guidance. Use the definitions in your current dashboard for platform metrics. The official references below help with checking; the Seller Academy link covers the US market.
- TikTok Shop: How to Use Product Analytics (US)
- TikTok for Business: About TikTok Shop Ads reporting time
Sources checked