The Monday call opens the same way more often than we would like. Our PPC lead says the hero SKU went out of stock on Thursday. The brand’s ops manager says yes, they knew on Tuesday, the PO slipped a week.
Between Tuesday and Thursday the account kept bidding at full strength into a listing that was about to go dark. Between Thursday and Monday it ran zero velocity into a competitive set that did not slow down. Nobody made a bad decision. The agency made decisions without a piece of information the brand had in a Slack channel the agency was not in.
After managing hundreds of brands on Amazon, the pattern we see most often is not a bad call by either side. It is a decision that was right on the information available and wrong on the information that existed. Working with an Amazon agency well is mostly a question of what reaches them before it lands in a report.
Why the agency is always the last to know
An agency’s inputs are almost all lagging indicators. Sessions, conversion, ACOS, rank, attributed sales that restate for a week after the fact. By the time an operational event shows up in any of those, it has already happened, and the account has been running on the old assumption for days.
The events that move those numbers live somewhere else. Inventory lives with ops. Price lives with finance or the founder. Packaging lives with the product team and the supplier. A DTC promo lives with the email person. None of those people think of the Amazon agency as someone they report to, because the agency “does Amazon” and this is not Amazon.
So the account is managed by people who see the effect and never the cause, and the post-mortem three weeks later is about ad efficiency, because that is where the reporting lives.
The nine events, what they cost, and when to flag them
These are the events that show up most in our takeover audits as the hidden cause of a number that got blamed on something else. For each one: what happens when the agency does not know, and how early it needs to hear.
1. Inventory is going to run short. A PO slips, a container sits in receiving, a SKU sells faster than forecast. If the agency does not know, bids stay at full strength into a listing that will be dark in four days, and when it goes dark the paid velocity that was holding rank stops with no plan for the re-buy. The reverse also costs money: a brand that quietly throttles by raising price to stretch cover, without telling the agency, produces a conversion drop that gets diagnosed as a creative problem. Flag it the moment cover on any top-20 SKU drops under four weeks, not when it hits zero.
2. A price change, list or MAP. Price moves conversion, conversion moves ACOS, and an ACOS move with no known cause gets a bid change. A brand that raises price 8 percent on a Wednesday and mentions it on the following Monday’s call has spent five days having its account optimised against a mystery. It also breaks deals: a Prime Big Deal Days submission made against the old price can run at a depth nobody approved. One week of notice, minimum, and before any deal submission closes.
3. A packaging, formula, count, or size change. This is the expensive one, because the damage is delayed. The moment the new units ship, every image on the listing describes a product that is no longer in the box. Returns start, “Customers say” begins collecting a theme about the difference, and the agency reads a slow conversion drift six weeks later. The trigger for telling the agency is not when the new units land. It is when the PO for them is placed, because creative takes weeks and an A+ resubmission clears a queue that runs long in Q4.
4. An off-Amazon promotion. A 25 percent code in a DTC email, a Costco road show, a Walmart rollback. Amazon’s fair-pricing behaviour can suppress the Buy Box on a listing priced above what a shopper can find elsewhere, and the agency will see a listing that stopped converting with no visible change to it. Two weeks of notice, and a decision on whether Amazon’s price moves in step.
5. A PR hit, a creator drop, or a TV spot. Branded search spikes. Campaigns capped for a normal Tuesday hit their daily budget by 11am. Worse, the new traffic types phrases nobody has seen before, and if a negative keyword list was built for old behaviour it blocks the exact terms the coverage created. Tell the agency the airdate the day you know it.
6. A retail launch. Going into Target or Walmart stores changes the comparison set, reprices the shelf, and shifts some Amazon demand to a physical aisle. It also creates a price parity question the brand has usually not decided. Six to eight weeks, because it changes what the agency should be defending.
7. A supplier or factory change. Different lot codes, sometimes a different UPC, and frequently a small quality variance that surfaces as a review cluster three months later. The agency needs this to read the reviews correctly. A sudden run of “quality has gone down” reviews on a listing with no other change is a supply event, not a listing event, and only one of those has a creative fix.
8. An internal approver is going to be unavailable. A vacation, a finance close, a budget freeze. The agency sends a dated ask, it sits, and the deal window or A+ deadline passes. The most common reason a Q4 creative brief ships late is a person who was on a plane the week it needed signing. Name the backup before the absence, not after.
9. Something you did in the account yourself. A VA edits a title. Ops uploads a flat file. The founder changes a coupon on a Sunday. The agency sees a number move on Monday with no cause, and spends a week investigating a change nobody logged. If anyone on the brand side has write access, every change they make needs a one-line note to the agency the same day. Better, it needs a row in a change log both sides can read.
What it costs on a real account
Take a $200K a month brand with a $60K a month hero SKU and $12K a month of ad spend behind it.
The PO slips and nobody tells the agency. Three days of bids at roughly $400 a day into a listing that cannot sell is $1,200 of spend with nothing behind it. That is the small number. The rank lost over the dark days gets bought back over four to eight weeks of paid support at an ACOS the brand would have refused in September, which on this SKU is another $3,000 to $5,000. Then the re-buy lands inside a peak window where CPCs run well above baseline.
Now add an unflagged price change the same month. Conversion drops two points, the agency reads it as efficiency, and cuts bids on the top campaigns in October. That is a decision made in the most valuable ten weeks of the year against a number that was explained by an email nobody sent.
None of this appears as a line item called “information the agency didn’t have.” It appears as a rough Q4 and a January conversation about whether the agency is any good.
The fix is a habit, not a tool
We have watched brands try to solve this with a shared project management board, and it fails because the people who hold the information do not live in the board. What works is smaller.
One shared operations calendar. A single document with every dated event the brand knows about: POs, expected landing dates, price changes, promos on any channel, launches, approver absences. The agency reads it weekly. It takes the brand ten minutes a week to maintain.
The “before” rule. The agency hears about a change before it happens, not after. Written down as a rule, because it is the opposite of how most brands operate by default.
One named owner on the brand side for outbound information. Not the founder. Somebody whose job includes telling the agency things. When that job belongs to everyone it belongs to nobody, and the stockout reaches the agency through a dashboard.
A five-line weekly note. Inventory position on the top 10, anything changing on price, anything landing from the supplier, anything happening off-Amazon, anyone unavailable. Five lines. Most weeks it is boring, which is the point.
The agency’s half. A good agency asks. If your agency has never asked about inventory cover, upcoming price moves, or what is happening on your other channels, that is a finding about the agency. The events above are known to every competent operator. An agency that waits for the dashboard is one that has decided the dashboard is its job.
Why this matters more in the next ten weeks
The number of these events per week roughly triples between now and December. Deal submissions close, inbound cutoffs pass, POs land or do not, promotions run on three channels on three calendars, approvers take time off, and the founder makes a Sunday-night coupon change because peak is stressful.
Every one of those is a decision the agency will make correctly or incorrectly depending on whether it knew. The calendar above costs an afternoon to build in September and nothing to maintain. In November it is the difference between an agency managing your account and an agency managing a version of your account that stopped existing on Tuesday.
FAQ
Isn’t it the agency’s job to monitor inventory?
Monitoring tells them cover is dropping. It does not tell them the PO slipped, the supplier is switching factories, or the brand plans to throttle by raising price. The agency can watch the number. Only the brand knows the cause.
We use Slack with our agency. Doesn’t that cover it?
Slack is where information goes to be seen once. A change made on a Wednesday and mentioned in a thread on Thursday gets scrolled past by Friday. The operations calendar is a standing document, not a conversation.
Our agency never asks about any of this. Should we be worried?
Yes. The events above are known to every experienced operator. An agency that does not ask about inventory, pricing, or your other channels is either assuming you will volunteer it or has decided those things are not its problem. Both cost you money.
What if the information is confidential, like a supplier change or a retail deal?
Give the agency the effect, not the detail. “Expect a new lot with slightly different packaging in October” does the job without naming the factory. “A retail launch is coming in Q1 that affects price” is enough to plan around.
How far ahead is far enough?
Far enough for the agency to act. A price change needs a week. A packaging change needs the creative lead time, which is weeks. A stockout needs the moment you know it is likely, not the moment it is certain. The rule is simple: if it changes what the agency should do, they need to know before they do the thing.
—
If you’re looking for a team that manages every lever — creative, advertising, and operations — Velocity Sellers works with brands doing $100K+/month on Amazon. Contact us for a free account audit.
