Before you treat chasing as a compliance problem, check whether it is really a capture problem. If a fact keeps having to be requested, one thing worth checking is whether it was ever recorded at the moment someone knew it. Automating the reminder makes the asking cheaper and the underlying design permanent.
If a piece of information has to be chased, the thing worth fixing is where it gets recorded, not who is slow to reply.
It is day eight of the close. There is a spreadsheet tab called ‘waiting on’, and it has the same three names on it that it had last month. One of them has already sent a figure that does not match the system, so now there is a second email asking which number is right, and nobody in the exchange is being difficult.
The waiting on list
Plenty of businesses keep a version of that tab under some other name. The chase list for timesheets, expenses, project status, stock counts, the CRM fields that never get filled in. The people on it are busy, the request landed on a day that was already full, and answering it means going back and reconstructing something they finished three weeks ago.
That last part is worth sitting with. Often, the person being chased knew the answer at the time. They knew how long the job ran while they were running it. They knew the count was short when they walked past the shelf. The information existed. It just never landed anywhere the business could reach later, so now somebody has to go and ask for it back.
What the chase is actually marking
A chase is a small piece of evidence about how the work was designed. It shows you a spot where the record of the work got separated from the doing of it. Something that was obvious to one person in the moment was turned into a question to be answered later, from memory, on somebody else’s schedule.
A request then behaves like a request. It competes with everything else in the inbox, it arrives without the context that made the answer easy, and it tends to get answered approximately. That is one way you end up with a figure that does not match the system and a second email trying to settle which one to believe.
This is not the only reason things get chased, and the others are worth ruling out first. Sometimes the fact genuinely does not exist until the period is over. Sometimes an approval has to clear before anyone can answer honestly. And sometimes the field is sitting right there and nobody realized they were the one expected to fill it, which has less to do with capture and more to do with nobody ever defining whose job it became.
Why the reminder is the easiest thing to buy
An automated chaser is close to a perfect demo. Polite nudges on a schedule, escalation when someone goes quiet, a clean summary of who is outstanding. It is cheap, it is quick to stand up, and it produces something visible in the first week. A reminder is also a rule rather than a judgment, which makes it a genuinely sound automation candidate on its own terms. That is exactly what makes it hard to refuse.
The trouble is what it quietly preserves. The friction of chasing is often what forces the harder question. Take enough of that friction away and the harder question can stop getting asked. A month later, the same names may still be outstanding. They are just being chased more politely, and the cost has moved from visible to invisible. It is the same trap as buying a demo of something you already pay for, except here you are paying to keep a design you would have changed if it had stayed annoying.
What to ask before you automate the follow-up
Take the chase list as it stands and ask three things about each item. Who knew this first. When did they know it. What were they doing at the moment they knew it.
The answers sort the list into groups. Some items have a clear moment where a person already had the fact in hand and was already in a system. Capture it there instead. That is often a small change: a field on a form someone already submits, a status that updates when a job is marked complete, a count entered at the shelf rather than remembered at the desk. Not an AI project, and sometimes not a purchase at all.
Some items have no such moment, because the answer truly does not exist until later. Those need asking for, and a reminder that does the asking is a fair use of the technology. The rest sit in between, where capture is possible but nobody has decided what should carry it and what stays a human step. That decision is the real work, and a nudging tool can make it easier to skip.
Chasing is cheap to smooth and expensive to keep. Before you buy anything that nudges people on your behalf, go through the recurring requests and find the ones asking someone to remember a fact the business could have written down the moment it was true. A tool that gets those answers back faster is still built on the assumption that they have to be asked for. The better question is not how to make people respond quicker. It is why this needs asking at all, when somebody already knew it three weeks ago.
See how we clarify the right path
See how we clarify the right path
Have something you’re trying to fix, build, or make easier?
About the author
Missy Ross
Missy Ross is the Founder and AI Architect of Vero Dawn, an AI architecture and solutions company that helps businesses see what could work better, determine the right solution, and bring it to life. Before founding Vero Dawn she spent 21 years in internal audit across banking and manufacturing, at Audit Director level.