KDP Ads Audit Trail: How to Investigate a Bid Change Before You Revert It
InteliAds Team
KDP Ads
You open your Amazon Ads campaign and see that orders slowed, a keyword bid looks lower than you expected, and you cannot say when it changed or why. The instinct is to type the old number back in and hope. A KDP ads audit trail is the record that lets you investigate first: what changed, when, at which level, and what else moved nearby. This guide covers how to build that record for Sponsored Products campaigns before you decide whether to revert.
Short answer: A KDP ads audit trail is a dated record of each bid, budget, target and rule change, paired with before and after performance for the same entity. Gather it before you revert, because reverting overwrites the current state you need to examine. Pin the change to a specific campaign, ad group and target, then compare equal windows around the timestamp. Check price, rank, seasonality and other ad changes as competing explanations. Then choose a full revert, a partial move, or no change, and log that decision too.
Why reverting a bid change too fast can erase the evidence
A bid is a single number that gets replaced. Once you type the old value back, the lower bid you were investigating is gone from the live campaign, and so is your ability to watch how it behaves for a few more days. If you did not write down the earlier value, the timestamp and the results around it, you have only memory to work from.
There is a second risk. A revert is itself a new change, and it lands on top of whatever else has moved since. Orders may have arrived at the lower bid, other targets may have been edited, and a scheduled rule may have run. Reverting without checking the current entity state can undo a correct fix along with a mistaken one, and it does not promise an exact reset of how the campaign was performing.
Bid-desk tip: treat the first hour after you notice a problem as evidence collection, not repair. Take screenshots or exports of the campaign and target view, note today's bid values, and only then decide.
Build a kdp ads audit trail from what you can actually see
Start with what your accounts expose, and be honest about the limits. The Amazon Ads console may show change history for some entities, but do not assume it is complete, that it names who made each change, or that it covers every field. Check what your own view displays for the campaign, and treat anything it lacks as a gap you fill yourself.
If you use InteliAds, keep the sources distinct. Amazon Ads is where the live bid actually sits. InteliAds is a KDP Sponsored Products manager connected through Amazon Ads OAuth, and its Rules run on your own IF/THEN conditions, scheduled hourly. History from a Rule applies to changes that Rule made, not to edits you or a teammate made directly in Amazon's console. BidBot gives separate bid recommendations, so note whether a change came from a rule, a recommendation you applied, or a manual edit.
Performance context improves when KDP Sync is linked, since it imports royalties, orders, KENP and pricing. InteliAds profit is royalties minus ad spend, so you can see whether a lower bid changed profit and not only ACOS. Without KDP Sync, record royalties from your KDP reports by hand.
Pin down the change: what moved, when, and at which level
Many "bid changes" turn out to be something else. An ad group default bid, an explicit keyword or product-target bid, and campaign bidding settings affect different parts of delivery. Check the value attached to the entity you are investigating. Confirm which level changed before you compare any numbers.
- Open the target view and note the current bid for each keyword or product target in the affected ad group.
- Check campaign-level settings for bidding strategy and placement adjustments, since these shift what you pay without touching the listed bid.
- Look for console history on the campaign if your view offers it, and record only what it shows.
- Review your InteliAds Rules if you use them, including when each last ran and what it was set to change.
- Ask anyone else with account access, such as an assistant or agency, what they edited this week.
Then write the change down in a consistent form. The values below are clearly hypothetical and only show the format.
| Logged at (hypothetical) | Entity ID (hypothetical) | Field changed | Before | After | Source |
|---|---|---|---|---|---|
| 2025-03-04 09:15 | HYP-TGT-0412 | Exact keyword bid | $0.62 | $0.38 | Manual edit |
| 2025-03-06 14:00 | HYP-CAMP-0107 | Top of search adjustment | +25% | +0% | Rule run |
A log like this separates two overlapping changes that would otherwise be blamed on one bid. In this hypothetical, the second entry landed two days after the first, so a before and after comparison around the first timestamp would be partly contaminated by the second.
Separate the bid effect from seasonality, price, and rank shifts
After you know what changed, test the alternatives. Compare equal-length windows on either side of the timestamp, ideally the same weekdays, and exclude the change day itself. Then check whether the book's price moved, whether a promotion started or ended, whether a holiday or release shifted demand, and whether organic rank or reviews changed.
Use an illustrative case. A hypothetical exact-match target drops from $0.62 to $0.38. In the seven days before, it logs $41.30 in spend and 9 orders. In the seven days after, it logs $24.10 and 4 orders. Spend fell about 42 percent, which fits a lower bid, but the orders fell too. If the book price rose by $2 during the same week, the order drop may belong partly to price. If the price stayed flat, the lower bid is a stronger suspect.
Do not read too much into a small count of orders. Four orders versus nine is a weak signal, and one more or fewer sale changes the story. Rules Dry Run in InteliAds previews the mutations a rule would make, not predicted sales, spend, ACOS or profit, so use it to check which entities a proposed rule would touch, never to forecast the result.
Once the evidence is gathered, match it to a decision. Read current entity state and any orders since the change before you act.
For the wider context on structuring Sponsored Products campaigns, read the KDP ads guide. To keep spend, ACOS and royalties visible across accounts while you investigate, the KDP advertising dashboard puts them in one place, and you can find more KDP ads guides on related campaign decisions.
Use one keyword and a consistent comparison window to turn the change record into a decision.
Run a worked check: before-and-after windows for one keyword
Take one keyword, one match type, one campaign. Compare two equal windows on either side of the logged change time, and leave out the change day itself. The numbers below are a labeled hypothetical: an exact-match keyword, "cozy mystery box set", in a Sponsored Products campaign for a $4.99 ebook with an assumed $3.49 royalty per order.
| Window | Bid | Clicks | Spend | Orders | Sales | ACOS | Royalties minus spend |
|---|---|---|---|---|---|---|---|
| 14 days before change | $0.55 | 48 | $26.40 | 6 | $29.94 | 88% | -$5.46 |
| 14 days after change | $0.80 | 51 | $37.74 | 5 | $24.95 | 151% | -$20.29 |
Clicks barely moved (48 to 51), so the higher bid did not coincide with meaningfully more clicks in this small sample. Average CPC rose to about $0.74 and orders slipped by one. Spend is the clearest change, up about $11.34. That points to the bid, but with this few orders one sale in either direction changes the picture, so treat it as a lead, not a verdict.
Record the change before you judge it
Use the timestamp, entity IDs and bid values in your change record to choose matching comparison windows.
Bid-desk tip: write the entry at the moment of the edit. A reason recalled three weeks later is usually a guess.
- Export or note clicks, spend, orders and sales for the keyword for both windows.
- Check the campaign for other edits inside the same span: budget, placements, new targets, negatives.
- Confirm the book's price and listing did not change in the same weeks, since that alters conversion regardless of bid.
- Calculate royalties minus spend, not ACOS alone, because a 151% ACOS only matters relative to what each order earns you.
Decide: keep, revert, or step the bid back partway
The check supports one of three outcomes. Pick based on what the evidence isolates, not on what feels safest.
- Keep: clicks and orders both rose enough that royalties minus spend improved or held, and nothing else changed.
- Revert: spend rose with no real gain in clicks or orders, as in the hypothetical above, and no other change competes as an explanation.
- Step back partway: results are mixed, or the sample is thin. Moving from $0.80 to roughly $0.65 limits the downside while you keep gathering data.
Before reapplying any old value, look at the keyword's current state: its present bid, whether it is still enabled, and whether new orders arrived since your window closed. An old number is not an exact reset, because the auction and your listing have moved on. Neither a revert nor a hold guarantees better results.
Where InteliAds fits
InteliAds is a Sponsored Products manager connected through Amazon Ads OAuth. When KDP Sync is linked, it imports royalties, orders, KENP and pricing, and profit is shown as royalties minus ad spend. That makes the last column of the table above easier to read in one KDP advertising dashboard, and across several accounts if you manage them.
You can write IF/THEN Rules that run on an hourly schedule, for example lowering a bid when spend passes a threshold with zero orders. Rules Dry Run previews the proposed mutations only. It does not predict sales, spend, ACOS or profit. The InteliAds Rules history records what InteliAds rules did and is separate from the Amazon Ads Console, so manual edits made inside Amazon still belong in your own log. BidBot is a separate feature that offers bid recommendations, and you decide whether to apply them. Targeting filters help you find the keywords worth checking first. Plans are on InteliAds pricing.
FAQ
Q: Does the Amazon Ads console show a full history of who changed a bid and when? A: Do not assume it gives you a complete, per-keyword record you can rely on for every edit. Check what change history your account actually exposes, and treat any gaps as normal. Your own log is the dependable source, along with the InteliAds Rules history for changes its rules made.
Q: How many days of data should I compare before and after a bid change? A: Use equal windows, commonly 7 to 14 days, and extend them for keywords with few clicks. Exclude the day of the change. If either window holds only a handful of orders, the comparison is weak whatever its length.
Q: What should I do if I can't remember changing the bid at all? A: Start from the data. Find the date spend or CPC shifted, then check your notes, team members with access, and any InteliAds rules that were active. If the source stays unknown, record it as unattributed and avoid assuming the bid was the cause.
Q: How can I tell whether a sales drop came from a bid change or from a book price or cover change? A: Line up the dates. If clicks held steady but orders fell, conversion is the suspect, which points to price or listing changes. If impressions and clicks dropped after a bid cut, the bid is more likely. With KDP Sync linked, imported pricing helps you see when the price moved.
Q: Is it better to revert a bid fully or move it halfway back? A: Revert fully when the evidence is clear and nothing else changed. Move partway when results are mixed or data is thin, then reassess after another equal window. Either way, check the current bid and recent orders first.
Q: How do I investigate a change when automatic campaigns or bid placement adjustments were also running? A: Separate what you can. Compare the same keyword or target alone, note when placement adjustments or automatic campaigns changed, and avoid blaming the bid for spend that shifted by placement. If several edits overlap, your conclusion is a hypothesis, so change one thing at a time afterward.
Q: What details belong in a simple change log for a KDP ad account? A: Timestamp with time zone, campaign and keyword IDs, field changed, before and after values, reason, who made it, whether it came from a manual edit or a rule, and a planned review date.
Q: When is a bid change too recent to judge fairly? A: When your after-window is shorter than your before-window, or when orders may still be attributed to clicks from the days just passed. A few days with a handful of clicks is too soon.
A reliable kdp ads audit trail is a timestamped record, an equal before-and-after comparison, and a deliberate keep, revert or partial step. For more background, read the KDP ads guide and more KDP ads guides. When you want one place to review spend and royalties, See InteliAds plans or Create an account. Essential has no trial. Review the current settings before enabling automated changes.
See InteliAds plans: compare Essential, Pro, and Publisher (Essential has no trial; Pro and Publisher include a 7-day trial). InteliAds is a Sponsored Products manager for KDP authors: royalty-aware profit when KDP Sync is linked, user-written IF/THEN rules with dry run, and targeting filters for keywords, product targets, and search terms. Create an account.