Automated weekly marketing summary

Start with the decision the summary should support

Choose the meeting or decision this report serves. A small service business may need to know which channels brought relevant enquiries and which follow-ups need attention. That calls for a different summary from a content team checking publication activity.

Limit the first report to measures people can define and act on. For each measure, write its source, owner, reporting date and exclusion rules. A contact-link click is evidence of an interaction with the link. It is not proof that a conversation happened or that the person became a customer.

Freeze the reporting window and record freshness

Pick a timezone and an inclusive date window. For example, report the previous complete Monday-to-Sunday week and compare it with the preceding Monday-to-Sunday week. Save those dates in the report so a later refresh cannot silently change what the reader is comparing.

Record when each source was retrieved and the latest complete date it supports. Different platforms process data at different speeds. Google Analytics explains that report data can change while processing completes and that attribution can be revised later. Set a cutoff that fits the sources you use.

If one source is late, show its last complete date and hold the affected comparison. Do not fill missing values with zero. An unavailable campaign export and a campaign with zero enquiries mean different things.

Join business records using a documented rule

A person may visit from search, return directly and later send a message. Avoid adding each platform attribution total together as if the platforms described different customers. Decide which source is authoritative for the count of genuine enquiries and sales.

Where your systems permit it, connect a recorded enquiry to its campaign or landing-page information using a stable enquiry ID. Keep unknown source as a valid category. Guessing a channel from an email address or a person’s wording creates confidence without evidence.

Use only the fields the summary needs. Aggregate team reporting can often use counts and categories without listing private messages or personal details. Keep access to the underlying enquiry records limited to the people who need to investigate them.

Use a worked comparison that explains the denominators

Illustrative example: a business records 24 genuine enquiries this week and 20 last week. That is an increase of four enquiries, or 20%. It classifies 9 this week and 10 last week as qualified. The qualification shares are 37.5% and 50%, a decrease of 12.5 percentage points.

The summary should show both observations. More enquiries alongside fewer qualified enquiries deserves a closer look at the campaign mix or the qualification process. The numbers alone do not establish that a specific advertisement caused the change.

If the previous value is zero, show the absolute change and mark the percentage comparison as not applicable. Include the sample sizes beside rates. A rate based on a handful of enquiries can move sharply after one new record.

Let the automation draft facts and the team add context

Generate a fixed opening: period covered, source status, key movements and outstanding questions. Rule-based text is enough for straightforward comparisons. If an AI tool drafts commentary, provide the checked aggregate table and require every numerical statement to match it.

Ask a reviewer to add context that is absent from the data: a campaign paused, tracking changed, a holiday shortened the week or a salesperson was unavailable. Keep explanations labelled as observations or hypotheses until supporting evidence exists.

Give each agreed action an owner and review date. A summary that ends with Investigate enquiry quality is less useful than one that names the person who will compare the two campaign groups before the next meeting.

Release a dated version and measure reporting effort

Send a dated approved summary to a maintained recipient list. Preserve the source timestamps and the version used. When an important source changes later, issue a clearly marked correction rather than silently replacing a file that people have already used.

Test the report with a missing source, zero baseline, duplicate enquiry and changed qualification rule. Confirm that it highlights the problem without inventing a trend. Compare preparation time and correction counts over similar weeks before claiming a benefit.

Bring your current weekly report and a list of its sources to Automations For Business. We can discuss how to collect the facts, check them and prepare the update around your existing tools. The downloadable brief below is a useful starting point for a WhatsApp conversation.

Your ready-to-use checklist

Make the next step clear.

Sources and notes

Sources checked on October 19, 2026.

• Google Analytics: Data freshness

  Processing delays and later changes can affect a weekly reporting cutoff.

• Google Analytics: Key event attribution models report

  Attribution comparisons depend on the selected model and reporting context.