Business superintelligence that acts.
Diagnose:
A drop in sales, broken into its causes, each with a dollar figure.
This is the difference between “sales are down 14%” and knowing that $17,000 of the drop is a promotion ending, $8,900 is a budget cut you chose, and $4,100 is a tracking tag that stopped reporting.
One question a week on the free account. Every account, every night, on the paid levels.
Why did sales drop in September?
Net sales were $182,400 against $212,900 in August, a drop of $30,500. Three things explain most of it.
Every figure links to its source. Illustrative account.
The six steps:
- 01
Frame the question
Which number moved, over which period, compared with what. Strategist repeats the question back before it starts, so the answer is to the question you actually asked.
- 02
Use the real revenue
Revenue comes from your store's real orders, not from what the ad platform claims. When the two disagree, both are shown and the gap is listed as a cause of its own.
- 03
Split the change
The change is split by channel, campaign, product, region and customer type. The splits that explain the most are kept. A cause has to account for real dollars to make the list.
- 04
Name the events
Promotions, budget changes, new creative, products out of stock, tracking gaps and competitor moves are read from the account's own change history and the store's calendar, and matched to the splits.
- 05
Rank and write
The causes are ranked by dollar effect and written in plain sentences a client or an owner can read, with every figure linked to its source.
- 06
Recommend, or refuse
A fix is suggested only when the data supports it and its effect can be predicted. When it cannot, the answer says so, and says what would settle it.
The answer it produces:
An example answer.
Why did sales drop in September?
Net sales were $182,400 in September against $212,900 in August, a drop of $30,500 (14%). Three things explain most of it.
First, the August promotion. The 20% code ran until 28 August and 38% of August orders used it. September had no code and 11% fewer orders, with average order value up $4. About $17,000 of the gap is the promotion ending, not a fall in demand.
Second, Meta prospecting spend was cut 22% on 9 September after the pacing alert, and new-customer orders from Meta fell 19% in the three weeks that followed. About $8,900 of the gap is here, and it was a decision, not a leak.
Third, GA4 conversions stopped reporting from 14 to 17 September while orders in Shopify continued. Reported revenue for those days is about $4,100 too low. The sales happened; the report did not see them.
What to do: keep the prospecting cut, because profit improved by $2,300 for the month. Fix the GA4 tag before the next report. Decide whether October gets a promotion or holds price. Last October held price and grew 6% on September.
Every figure above comes straight from the account and the store. AI wrote the sentences, not the numbers. Illustrative account: invented to show the format.
Business superintelligence that acts.
Free to start. One Adella account for every app.