How is a monthly client report prepared today?
It starts with several tabs open. The data lives in three different places: the invoicing software, the tool where projects are managed and a spreadsheet that somebody updates by hand. Whoever prepares the report exports, copies, pastes, checks that the numbers add up and arranges them in the usual format. Then they write a comment and send the document.
With a few clients, it is a short job. With more, that person spends a large part of their time on work that barely changes from one client to the next. And the report turns up late if someone is off sick or if one of the sources had not been updated.
There is also a risk that is easy to miss. When you copy and paste by hand, a number can be left over from a previous month or get mixed up between clients. The recipient may spot it, and then the doubt falls on the whole report.
Which part of the report can be automated?
The part that consists of collecting and arranging data. An automated process can do the following every month, at the moment you choose:
- Read the data from each source, provided it has an API or allows an export.
- Check that all the expected data has arrived.
- Work out the figures you always calculate and put them in your usual format.
- Prepare a draft of the document for each client.
- Leave it in a folder, or send it to whoever has to review it.
If something fails, for instance a source does not respond, the system tries again and alerts a person. That client's report waits until someone confirms that it can carry on.
An AI tool can also suggest a first piece of text that describes the figures in plain language, such as "visits went up compared with the previous month." It is a starting point. Someone on your team reads it, corrects it and decides what the client is told.
What should stay with a person?
The judgement. A report also says what has happened and what will be done next, and that depends on what you know about the client: what worries them and what explanation they need when a figure drops.
That part is done by a person on your team. If the report goes straight to the client without anyone reading it, a mistake turns into a question of trust. A sensible split is that the system prepares the draft and a person reviews it and adds the commentary before it is sent.
What should you decide before you start?
Four decisions separate a useful automatic report from one that nobody wants to use.
- What the report contains. Many reports pick up extra sections over the years. It is a good moment to ask which ones the client actually reads and which stay out of habit.
- Where each figure comes from. One source for each piece of data. If the same number can come from two places and sometimes they disagree, you need to decide which one wins before automating anything.
- Who reviews it, and what they can change. One named person, who can edit the draft before it goes out.
- Which data goes in. Decide which client and project data the process uses, and which stays out, before connecting any source.
The second decision can expose differences. Doing the exercise may show that two people calculate the same indicator in different ways.
How does it change the work of the person preparing the report?
They stop spending their attention on copying and pasting and move to reviewing. They open a draft that is already put together, check the figures that look odd, add the commentary and send it. The time that used to go on exporting is used to think about what the client wants to know.
For the client, the report always arrives in the same format and at the same moment. That builds confidence, because they know what they will find and where.
For you, the gain is that the report stops depending on a single person. If that person is away, someone else can open the draft and finish it.
How to get started
Pick one report you already produce every month and note where each piece of data comes from, who prepares it and what commentary they add. With that page, you can see what repeats and what needs judgement.
If you want to check whether your case is a fit, get in touch. We work at a fixed price per project, we hand everything over documented, and the system is yours. If your reports are not a good candidate for automation, we will say so. You can see more on the automation page.