What we found
The part of the Genesco build that pays again at every close is the memory it keeps between runs. The client is Genesco Sports Enterprises. Every month, the finance team rebuilt job-profitability numbers by hand from Sage 50 exports. One of those steps was re-keying prior-month profit figures, so the new month started from the right place. We built a two-pass web app for that close. It remembers each month's numbers so the next month starts pre-filled. A month-end reconciliation that took the CFO about 4 days by hand now runs in about 60 seconds.
The figures here are Genesco's and they are public with their permission. They describe one company and one close. An earlier Field Note has the before and the after of that reconciliation, and judgment calls stay with finance has the line between the app and the team. This piece is about the state the app carries from one month into the next, and what that one design choice removed.
The re-keying step
Re-keying prior-month profit figures existed because those numbers lived in last month's file. Someone opened the finished workbook, found the figures the team had agreed on, typed them into the new month's sheet, and then checked that the totals tied.
Two things made that step worse than it looks. It grew with the business, because more jobs meant more lines to carry across. And it sat at the front of the close, where a mistyped figure travels through everything after it. A number carried over wrong is hard to spot later, because every downstream total agrees with it.
When we mapped the close with the CFO, this step had no owner's decision in it. Nobody weighed anything. A person was being used as the connection between two files.
The two-pass flow
The app reads the Sage exports, computes the job totals, and pre-fills the figures that follow a written rule. The first pass does the arithmetic and the carry-forwards, then hands back a clean workbook. The finance team works in the middle, in a familiar spreadsheet, on the short list of items that need a decision. The second pass takes that reviewed workbook and produces the final reconciled master.
The memory sits at the end of the second pass. When a month is finished, the app keeps the agreed figures for that month. The next month's first pass reads them as the opening position and fills them in before anyone touches the file. Each run also leaves a record of what went in and what came out, so a figure can be traced back to the month it was agreed in.
The memory here is stored figures. The app does not learn anything from them, and it does not estimate. It holds what the finance team signed off on last month and puts it where that number belongs this month. The arithmetic around it runs on rules written in code, so the same files produce the same answer every time.
Month two
Month two starts with last month's agreed figures already in place. The team uploads the Sage exports, reviews the short list in the workbook, and downloads the master. The opening position is already there, and the reviewer's first act of the month is a decision rather than a transcription.
The setup cost lives in month one. Writing down the rules took a line-by-line pass through a real month with the CFO: which figures follow a written rule, which ones depend on context, and what the carry-forward is for each job. That work happened once. Every close after it reuses the same written rules and the same stored figures.
Keeping the figures also made re-runs cheap. When a late number arrives, the team runs the pass again and reviews a fresh workbook, and the carried figures do not have to be typed in a second time. Back when a close cost days, a re-run was hard to justify, so a late number waited for the next cycle. That is a change in what the team can do, alongside the change in how long the close takes.
Limits
This design works because a finished month at Genesco has an agreed set of figures. The app can only carry forward numbers that someone has signed off on. A close that reopens prior months as a matter of routine needs a rule about which version is the one to carry, and that rule has to be written before any of this helps.
The other limit is the same one we name every time. These numbers belong to one company and one process. Genesco's close had clear inputs, clear outputs, a large amount of manual work, and a rule set that could be written down with the people who run it, which is what made it a good first process. We wrote about picking one in which process to automate first. A close where judgment enters at every step compresses far less.
A build like this also needs an owner on the client's side. At Genesco that is the CFO, who used it, told us what to adjust, and got the adjustments. Mark Tatum, Genesco's CFO, described it this way: "Every month our team burned hours rebuilding job-profitability numbers from Sage by hand. Plainpath built us a tool that does the heavy lifting and leaves us just the judgment calls. It has saved us real time and headaches, and they were quick to adjust it when we asked." A build with nobody asking for changes drifts away from the work, which is the pattern in why most AI pilots fail.
The plain take
Write down your own close in the order it happens and mark every point where a person reads one file to fill in another. Those points are carry-forwards, and a carry-forward is a figure software can hold between runs once someone writes down the rule for it. Our free audit asks about your work and scores it, and the questions it asks are here. Discovery is where a person checks that and writes the build plan.
Frequently asked questions
What does it mean for an app to remember last month's numbers?
At Genesco Sports Enterprises, the app stores the figures the finance team agreed on at the end of each close. The next month's first pass reads those stored figures and fills them in before anyone opens the workbook. It holds signed-off numbers and does not estimate or learn from them.
Why was re-keying prior-month figures a problem?
The step repeated work that had already been done, it grew as the number of jobs grew, and it sat at the front of the close. A figure typed wrong at the start travels through every total after it, which makes it hard to catch later.
Does the finance team still make the decisions?
Yes. The app pre-fills the figures that follow a written rule and hands back a clean workbook for the judgment calls. A reviewer who disagrees changes the number in the workbook, and the second pass takes that change as given.
Will our month-end close compress the same way?
We cannot promise that. The Genesco figures cover one company and one process, and that close had clear inputs, clear outputs, heavy manual work, and rules that could be written down. A close where judgment enters at every step compresses far less.
How do we find this step in our own process?
Write the work down in the order it happens and mark every point where a person reads one file to fill in another. Our free audit walks through that, and Discovery is where a person checks it and writes the build plan.
The plain take
Mark every point in your close where a person reads one file to fill in another. Those points are carry-forwards, and a carry-forward is a figure software can hold between runs once someone writes down the rule for it. Our free audit asks about your work and scores it, and Discovery is where a person checks that answer and writes the build plan.
If one of your processes looks like this, the free audit at /compass is where to start.