Set both date ranges in custom mode. The two pickers are completely independent — ARM does not work the prior range out from the present one, and it does not check that the two are the same length or a year apart. Leave the prior-year picker on its default and you can end up comparing three months against three weeks, with every percentage on the row worked out from those two mismatched periods and nothing on screen to warn you. In the example the present range is 06-01-2026 to 08-31-2026 and the prior range is 09-01-2025 to 09-24-2025, which is why the changes look so severe.
Weekly mode compares to 52 weeks back, not to the same date last year. Pick Saturday 07-04-2026 and the prior-year column shows 07-05-2025 — 364 days earlier, not 365. That is deliberate: it keeps a Saturday against a Saturday so you are comparing a full trading week with a full trading week.
The row begins with City and State, then the Present Year Amounts block: Week End/Custom Date, Gross Sales (Minus Sales Tax and Refunds), Car Count, Net Refund Sales, Net Final Sales and Ticket Average.
The heading on the main money column is not accurate, and it matters. Gross Sales (Minus Sales Tax and Refunds) has your sales tax taken out but the refunds are still in it. The refunds come off one column further along, in Net Final Sales. In the example the column reads $5,541.85, refunds were $1,333.69, and Net Final Sales is $4,208.16 — the subtraction has already been done for you. Do not take the refund off again.
Net Final Sales is the figure to trust for the period. It is the money you kept: the column before it, less Net Refund Sales.
Ticket Average does not use it. The average is worked out from Gross Sales (Minus Sales Tax and Refunds) divided by Car Count — in the example $5,541.85 ÷ 12 = $461.82 — so on a report designed to net off refunds, the per-ticket figure ignores them. The true after-refund average is $4,208.16 ÷ 12 = $350.68. Work it out yourself from Net Final Sales if the refund figure is anything but small.
Then comes Prior Year Amounts, with the same Week End/Custom Date, Gross Sales (Minus Sales Tax and Refunds), Car Count and Ticket Average for the comparison period; three % Change columns — In Net Sales, In Car Count and In Ticket Avg; a second Net Refund Sales and Net Final Sales pair; and finally True Present Year Net Sales.
The second Net Refund Sales and Net Final Sales columns are the prior year's, even though nothing above them says so — the heading strip over those two is blank, while the identical pair earlier in the row sits under Present Year Amounts. Count across from the left if you are not certain which pair you are reading.
The three % Change columns are worked out before refunds. Each compares the present year's Gross Sales (Minus Sales Tax and Refunds) against the prior year's, not the Net Final Sales figures. In the example In Net Sales shows −93.07%, which is $5,541.85 against $79,996.08; comparing what you actually kept — $4,208.16 against $79,937.09 — gives −94.74%. On a small refund the difference is slight; on a heavy refund month it is not.
True Present Year Net Sales is a different figure again. It is the net sales ARM stores against the invoices themselves, and it will not usually match the Gross Sales (Minus Sales Tax and Refunds) column at the start of the row — in the example $6,575.56 against $5,541.85. The two are built differently: one is taken straight from the invoice record, the other is worked out from gross sales less tax. Treat the start of the row as this report's own measure and True Present Year Net Sales as ARM's stored figure, and do not expect them to agree.
The printout carries all eighteen columns and matches the screen. Click PRINT and the PDF opens in a new tab, headed with your store details and Prior Year Comparison After Refund Report followed by your present-year dates. It is a tighter layout than the screen — the headings wrap onto several lines to fit the row across one page — but every figure on it is the figure you were looking at.
A note on the three reports in this family. Prior Year Comparison Report gives the same comparison without any refund columns; Net sales/Car count/Ticket average Prior Year Comparison Report adds a cars-per-day block instead. This one is the only one of the three that shows what you refunded, so use it whenever refunds matter and one of the others when they do not.
