Sorting emails, moving form data into the CRM, writing the first draft of a report, answering the same questions over and over: in any marketing team, a good chunk of the week goes on repetitive tasks that need consistency rather than creativity. Combining Make (to connect and automate tools) with Claude (to understand and write text) takes a lot of that work off your plate.
In this article I explain how the combination works, which automations are worth building first and how much time you can realistically save.
Make orchestrates, Claude thinks
Make is a visual platform where you connect apps (email, spreadsheets, CRM, WordPress, social networks, Slack…) through «scenarios»: a trigger that starts the process and a series of steps that run on their own. It has an official Anthropic Claude module, so you can add an AI step in the middle of any flow without coding.
The division of roles is simple: Make moves the data between tools and decides when to do each thing; Claude interprets and writes: it summarises, classifies, extracts data from messy text, drafts in your brand’s tone or decides which category a message belongs to. Without Claude, Make only repeats fixed rules; without Make, Claude only works when someone writes to it.
Five marketing automations that work
- Lead classification and response. A web form triggers the scenario; Claude reads the message and detects interest and urgency, Make creates the contact in the CRM with that tag and prepares a personalised draft reply.
- Content drafts for blog and social. From a spreadsheet row with the topic and keywords, Claude writes a first draft and Make saves it as a draft post in WordPress, ready to review.
- Content repurposing. Every published article automatically becomes a newsletter summary, social posts and several ad copy variants.
- Weekly reports in plain language. Make pulls analytics and campaign data, and Claude summarises it into a text covering what’s up, what’s down and what to check, which lands in the team channel every Monday.
- Review and mention management. When a new review comes in, Claude classifies the sentiment, suggests a reply in the brand’s tone and alerts a person if the case is sensitive.
In every case, a person should review the result before it reaches the customer or gets published: the automation prepares, the team decides.
How much time you save
A small team that builds the five automations above can win back around 15 hours a month, almost 75% of the time it currently spends on those tasks. This is an illustrative example with these assumptions: 40 leads a month, 4 articles a month, 4 weekly reports and 30 new reviews; the times are estimates, not measurements, and yours will vary.
| Automation | Before (min/month) | After, with human review (min/month) | Saved (min/month) |
|---|---|---|---|
| Leads: classify and prepare reply | 320 | 80 | 240 |
| First draft of articles | 360 | 120 | 240 |
| Content repurposing | 180 | 40 | 140 |
| Weekly report | 240 | 40 | 200 |
| Reviews and mentions | 150 | 45 | 105 |
| Total | 1,250 | 325 | 925 |
If you work out your own, multiply each task by its real frequency and always subtract review time. The saving isn’t only in hours: that time gets reinvested in strategy, customer relationships and creativity, which is what an automation can’t do for you.
How to start and mistakes to avoid
Start with a single task: the one that repeats most and carries the least risk. Sketch it on paper, build it in Make with a Claude module, test it with real data and launch it with human review. Once it works, add the next one.
- Clear instructions for Claude. Define the brand tone, the output format and what to do when in doubt; a good prompt prevents most failures.
- Review before publishing or sending. Results should arrive as drafts or in an approval channel, not straight to the customer.
- Be careful with personal data. Don’t send the AI more information than needed and check each tool’s privacy policy.
- Keep an eye on costs. Make and Claude are billed by usage: watch the number of operations and the length of the texts.
- Review your scenarios from time to time, because apps change and a flow that works today may break tomorrow.
If you’ve automated something similar in your team, I’d love to hear which process you started with: get in touch.
