Getting the files

Reporting the week is mostly a drag-and-drop. This is where each file comes from, what it fills in, and what no export will ever tell us.

The week is due Sunday evening at the latest — but nothing about it needs a weekend. Friday before you log off is the better slot, and the form fills in the same figures either way.

Before anything else

The files never leave your laptop. The entry form reads them in the browser, counts the rows that fall in the week being reported, shows you the figures, and only writes them into the form when you press the button. Nothing is uploaded, and no tool is connected to your LinkedIn account — which is deliberate: browser automation tools get accounts restricted, and a restricted account costs more than ten minutes of your week.

Every figure lands as a suggestion. Read it before you apply it, and correct anything that looks wrong — the number you submit is yours, not the file's.

Growth and outreach

Teodora, George, Jeremy. Two sources cover six of your fields.

LinkedIn — your own data archive

Ask a day ahead
  1. On desktop, open Me → Settings & Privacy.
  2. Go to Data privacy → Get a copy of your data.
  3. Choose Download larger data archive — not the fast file. The fast one only carries connections; we need invitations and messages too.
  4. Press Request archive and type your password. LinkedIn emails a link when it is ready — usually within a few hours, occasionally up to 24, which is the only reason this one needs planning: ask for it the day before you intend to report. Thursday if you file on Friday.
  5. Open the email, download the zip, unzip it. Drop Invitations.csv, Connections.csv and messages.csv onto the entry form.
What it fills
Invitations.csv → LinkedIn requests sent. Outgoing rows only, dated inside the week. Connections.csv → requests accepted, read off the Connected On date. messages.csv → replies from accepted, and a first guess at conversations engaged in.

Two things to know. Acceptances often belong to requests you sent weeks ago — that is expected, and why the chain is read on month-to-date totals. And the reply count works by telling your messages apart from theirs using your first name as it appears on the roster; if your LinkedIn display name is different, the file will say so instead of guessing.

Calendly — meetings booked and held

  1. Open Scheduled events in Calendly.
  2. Set the range to the week you are reporting, and leave the status filter on all events so cancellations come through too.
  3. Press Export and choose CSV.
  4. Drop the file on the form.
What it fills Meetings booked — every event created inside the week, including ones later cancelled, because it was still booked. Meetings held — events that started before today and were not cancelled.

A meeting booked this week for next month counts as booked now and held later, which is right. A no-show still counts as held unless you cancel it in Calendly — so cancel no-shows there and the show-up rate stays honest by itself. If you use Google Calendar or HubSpot instead, export to CSV from those and the form will read whatever date and cancellation columns it finds.

Email — HubSpot or Twenty

Drop the file in

Both tools can hand you a CSV of the week's email activity, and the form reads either. What it needs is one row per email with a date and a direction or status — it looks for a date column, then for words like sent, replied, bounced or delivered, and counts the rows that fall in the week.

HubSpot

  1. Sales email activity sits in the Marketing → Email list for sequences, and in Contacts → Activities for one-to-one sends. Sequences are the usual case for us.
  2. Open Automation → Sequences, then the Analyse tab. Set the date range to the week and the breakdown to by sender so you only carry your own sends.
  3. Press the export icon, choose CSV, and HubSpot emails you the file. Keep hubspot in the filename.
  4. Alternatively, a saved Activity list export works too: filter activity type to email, owner to you, date to the week, then Export view. Include the columns for activity date, activity type and email status.

Twenty

  1. Open the Messages object — or whichever object your email sync writes into, if it has been renamed.
  2. Add a filter on the date field for the week, and a second on the owner or sender for you.
  3. From the record-table menu, choose Export records → CSV. Twenty exports the visible columns, so put date, direction and status in the view before you export. Keep twenty in the filename.
  4. Twenty is open-source and its API is GraphQL, so once we are settled on it this step can be automated properly and the numbers appear without anybody exporting anything. Worth doing only after the tool choice is final.
What it fills Emails sent, email replies, positive email replies where a status or stage names them, and the bounce count behind the bounce rate.

Whichever tool wins, send me one real export and I will pin the column names exactly — until then the form matches on column headings, tells you in plain words what it found, and leaves anything it could not find for you to type. Neither tool can judge a positive reply, so check that one yourself in every case.

Creative content

Frans. The drop zone is on your form too, and it works on platform exports rather than a LinkedIn archive. One rule matters more than the steps: keep the platform name in the filename — linkedin, instagram, facebook, bluesky — because that is how each file is assigned to a platform count.

LinkedIn page

  1. Open the ChatTutor page as an admin, then Analytics → Content.
  2. Set the date range to the week and press Export.
  3. LinkedIn hands you an .xls. Open it, then Save as → CSV, keeping the word linkedin in the name. Use the sheet with one row per post, not the daily-totals sheet.

Instagram and Facebook

  1. In Meta Business Suite, open Insights → Content.
  2. Pick the week, then Export data → CSV, one export per platform.
  3. Name them so instagram and facebook appear in the filenames.

YouTube, if a video lives there

  1. YouTube Studio → Analytics → Advanced mode.
  2. Set the week, then the download icon → Comma-separated values.

Sessions and trials

  1. In GA4 → Reports → Acquisition, set the week and share or export the report as CSV.
  2. Any column called sessions is summed. Trials are picked up if a column names them; otherwise type the number.

Bluesky

There is no analytics export. Count the pieces you published and type it.

What these fill Pieces published per platform, impressions, engagements, video views, sessions and — when a column names them — trial signups.

What stays typed, and why

Positive replies. Somebody has to read the message and decide whether there is real interest. A classifier would produce a number nobody trusts, which is worse than no number. Conversations engaged in. The form makes a guess from the message file. Check it — a thread can have four messages and no conversation in it. Second conversations. Whether a first meeting genuinely led anywhere is a judgement about the institution, not a calendar fact. The buyer split. Which of the three segments a meeting belongs to is something only the person in the meeting knows. Videos and stories delivered, on time or not. Published is not the same as delivered on time, and only Frans knows which it was. Prospects added, warnings received. Both live in your own list and inbox rather than in any export. Meetings from other sources. Logged one by one on the form — where it came from, which institution, which buyer, and whether it has happened yet. A Calendly export cannot tell a referral from a cold reply, and the difference is the point.

These are also the smallest counts of the week, so the typing costs a minute. The ones that are tedious — requests, acceptances, replies, meetings, impressions — are the ones now handled for you.

The routine

Sunday evening is the deadline, not the appointment. File it Friday afternoon and your weekend stays yours — the week's counts are complete by then in all but the rarest case, and a Friday filing can be corrected on Monday if something lands late.

  1. The day before you file. Request the LinkedIn archive — growth roles only, and the one step that needs a head start.
  2. When you file. Export Calendly and your email tool for the week, or your platform files if you hold content.
  3. Open the entry form, unlock with your name, drop the files in.
  4. Read the figures. Apply them. Fill in the handful the files cannot know.
  5. Check the chain panel at the foot of the form — it names the first link that broke, which is what Monday is about.
  6. Anything that lands after you file goes in next week. Do not hold the week open for it.
  7. Send to the sheet, or copy the JSON and commit it.

If an export is missing or a site has moved its buttons, type the week by hand and say so on Monday. A typed week is worth more than a skipped one, and a week filed late is worth more than a weekend spent on it.

ChatTutor Human Thinking · Copenhagen