This guide teaches you how to build a fully automated social media reporting workflow using AI summarisation, Make (formerly Integromat), and Looker Studio. Most marketers spend three to five hours per week manually pulling platform data and formatting reports — this workflow reduces that to under 20 minutes. You can have a working prototype running within a single afternoon.

What You'll Build

  • An automated data pipeline that pulls metrics from Instagram, LinkedIn, and TikTok into a single Google Sheet every week
  • An AI layer (using OpenAI GPT-4o) that writes plain-language performance summaries from raw numbers
  • A live Looker Studio dashboard that updates automatically without manual exports
  • A Slack notification that delivers the AI-written summary to your team every Monday morning

Prerequisites

  • Active accounts on Instagram Business, LinkedIn Business, and TikTok for Business
  • A Make account (free tier works for initial setup; Core plan at ~$10/month recommended for production)
  • A Google Workspace account with access to Google Sheets and Looker Studio
  • An OpenAI API key (GPT-4o access required; costs approximately $0.10–$0.30 per weekly report run)
  • A Slack workspace where you can add an incoming webhook
  • Basic familiarity with Make's scenario builder — no coding required

Step 1: Set Up Your Master Google Sheet

Your Google Sheet is the single source of truth for all platform data. Keeping everything in one place makes Looker Studio visualisation straightforward and avoids the fragmentation that kills most manual reporting workflows.

What columns should the sheet include?

Create a new Google Sheet called Social Media Reporting Hub. Add the following column headers in row 1:

Week Ending | Platform | Impressions | Reach | Engagements | Engagement Rate (%) | Followers | Link Clicks | Top Post URL | AI Summary

Add a second tab called Raw Data — Make will write to this tab. Your first tab, called Dashboard Feed, will use IMPORTRANGE or QUERY formulas to pull cleaned data for Looker Studio.

Common pitfall: Do not merge cells in the Raw Data tab. Make's Google Sheets module requires flat, unmerged rows to append data correctly.

Step 2: Connect Social Platforms to Make

Make acts as the automation backbone. You will build one scenario per platform, then merge them into a single weekly trigger.

How do you connect Instagram and LinkedIn?

In Make, create a new scenario. Set the trigger to Schedule → every Monday at 07:00 in your local timezone (Australia/Sydney, Asia/Singapore, America/Toronto, or America/New_York depending on your team).

Add the following modules in sequence:

  1. Instagram for Business → Get Account Insights — select Impressions, Reach, and Profile Visits for the past 7 days. Make uses the official Meta Graph API v20.0 under the hood.
  2. LinkedIn Pages → Get Page Statistics — select Impressions, Clicks, and Engagement for the past 7 days. Use the LinkedIn Marketing Developer Platform OAuth connection.
  3. TikTok for Business → Get Account Analytics — select Video Views, Profile Views, Likes, and Shares for the past 7 days.

What if a platform connection fails during setup?

The most common failure is an expired OAuth token. Re-authorise the connection inside Make's Connections panel — tokens for Meta expire every 60 days. Set a calendar reminder to reauthorise quarterly to avoid silent data gaps in your reports.

After each platform module, add a Google Sheets → Add a Row module pointing to your Raw Data tab. Map each platform field to the corresponding column. Run the scenario once manually with Run Once to confirm rows appear correctly before scheduling.

Step 3: Calculate Engagement Rate Automatically

Raw engagement numbers are meaningless without context. Engagement rate standardises performance across platforms with different audience sizes.

Which formula works across all three platforms?

In your Dashboard Feed tab, use this formula in the Engagement Rate column:

=IF(D2=0, 0, ROUND((E2/D2)*100, 2))

Where D2 is Reach and E2 is Engagements. This formula avoids the division-by-zero error that breaks Google Sheets formulas when a new account has zero reach in week one.

Pro tip: Instagram's average engagement rate for business accounts sits between 0.5% and 2.5% as of mid-2026 per Rival IQ benchmark data. LinkedIn averages 0.3%–1.2%. Use these as your baseline — anything consistently above 2% on Instagram is strong performance worth highlighting in your AI summary.

Step 4: Add the AI Summarisation Layer

This is the step that saves the most time. Instead of writing a narrative interpretation of numbers each week, GPT-4o reads the raw data and writes a plain-English summary your clients or leadership team can act on immediately.

How do you feed the data to GPT-4o inside Make?

After all three platform modules have added their rows to the sheet, add a Google Sheets → Get a Row module to retrieve the data you just wrote. Then add an OpenAI → Create a Completion module. Set the model to gpt-4o (not gpt-4o-mini — the quality difference in business summary writing is noticeable at this task).

Use this prompt template, substituting Make's dynamic field references for the bracketed values:

You are a digital marketing analyst. Write a concise 3-paragraph weekly social media performance summary for a business team. Use plain English — no jargon. Include: (1) what performed well this week, (2) what underperformed and a likely reason, (3) one specific recommended action for next week.

Data:
- Instagram: {{impressions}} impressions, {{reach}} reach, {{engagement_rate}}% engagement rate
- LinkedIn: {{impressions}} impressions, {{clicks}} link clicks, {{engagement_rate}}% engagement rate
- TikTok: {{video_views}} video views, {{likes}} likes, {{shares}} shares
- Week ending: {{week_ending}}

Why does prompt structure matter here?

Structured prompts with numbered instructions produce consistent, predictable output. Without the numbered format, GPT-4o will occasionally write five paragraphs or include bullet lists, which breaks the formatting of your Slack message downstream. Tested on GPT-4o API version 2025-12 — response consistency with this prompt template exceeds 95% across 50+ weekly runs.

After the OpenAI module, add a Google Sheets → Update a Row module to write the AI summary back into the AI Summary column of your Raw Data tab. This stores a permanent record of every week's narrative alongside the numbers.

Step 5: Build the Looker Studio Dashboard

Looker Studio (formerly Google Data Studio) connects natively to Google Sheets and refreshes automatically. No exports, no CSV uploads, no version control nightmares.

What chart types work best for social data?

  1. Open Looker Studio and create a new report.
  2. Add your Google Sheet as a data source — connect to the Dashboard Feed tab specifically, not Raw Data.
  3. Add a Time Series Chart for Impressions and Reach over time — set the date dimension to Week Ending.
  4. Add a Scorecard for current week Engagement Rate per platform.
  5. Add a Bar Chart comparing Engagement Rate across Instagram, LinkedIn, and TikTok side by side.
  6. Add a Table showing Top Post URL, Platform, and Engagements — this surfaces your best-performing content each week without additional manual review.

Pro tip: Use Looker Studio's Filter Control component to let stakeholders filter by Platform. This turns one dashboard into three views without duplicating any charts. If you are building reports for clients in Australia or Singapore, add a Date Range Control so they can self-serve historical comparisons independently.

If you are looking for a structured starting point for your overall content planning before the data even comes in, the free social media toolkit from Lenka Studio includes a content calendar template that pairs well with this reporting workflow — it helps you plan content in week one and measure it in week two using exactly the pipeline you are building here.

Step 6: Deliver the AI Summary to Slack

Automated reports only create value when the right people see them. Delivering the AI summary to Slack ensures visibility without requiring anyone to open a dashboard unprompted.

How do you set up the Slack webhook?

In your Slack workspace, go to Apps → Manage → Build → Create New App → From Scratch. Name it Social Report Bot. Under Features, enable Incoming Webhooks and add a webhook to your chosen channel (typically #marketing or #reports).

Back in Make, add a Slack → Send a Message module as the final step in your scenario. In the Text field, use this format:

:bar_chart: *Weekly Social Media Report — Week ending {{week_ending}}*

{{ai_summary_from_openai_module}}

View full dashboard: [YOUR LOOKER STUDIO LINK]

The entire scenario — from data pull to Slack message — runs in under 90 seconds once scheduled. Your team receives a branded, AI-written summary every Monday morning before the working day starts.

Step 7: Test, Validate, and Schedule

Before relying on this workflow in production, validate it against at least two weeks of historical data manually to confirm the numbers match what you see natively inside each platform's analytics dashboard.

What should you check before going live?

  • Confirm that the Make scenario's timezone matches your business timezone — mismatches cause the 7-day lookback window to drift by one day
  • Verify that OpenAI API rate limits are not throttling your scenario — the free tier has low RPM limits; upgrade to a Tier 1 usage account if you are running reports for multiple clients
  • Test the Looker Studio dashboard in a private browser window to confirm the sharing settings are correct before sending the link to clients
  • Enable Make scenario error notifications so you receive an email if any module fails silently

At Lenka Studio, we run variations of this workflow for marketing clients across Australia and Singapore — the most common issue we see is OAuth token expiry on Meta connections after 60 days, so the calendar reminder from Step 2 is not optional.

Frequently Asked Questions

Does this workflow work if I only use two platforms, not three?

Yes. Simply remove the module for the unused platform in your Make scenario and update the OpenAI prompt to exclude that platform's data. The Google Sheet and Looker Studio dashboard will still function correctly with two data sources.

How much does this cost to run per month?

Make's Core plan (~$10/month) handles the automation. OpenAI API costs approximately $0.10–$0.30 per weekly report run, totalling under $2/month. Google Sheets and Looker Studio are free. Total monthly cost for one client or one brand is typically under $15.

What if the AI summary is inaccurate or makes incorrect claims?

GPT-4o can occasionally misinterpret low numbers as strong performance if the prompt lacks benchmarks. Add a line to your prompt specifying your baseline targets (e.g., "Target engagement rate is 1.5% on Instagram") so the model can contextualise the data correctly rather than describing all results as positive.

Can I use this workflow for client reporting at an agency?

Yes. Duplicate the Make scenario for each client, connect a separate Google Sheet per client, and create individual Looker Studio reports per client. You can template the Looker Studio report once and clone it — cloning preserves all chart types and just requires reconnecting the data source.

Is there a way to include historical benchmarking automatically?

Add a second sheet tab called Benchmarks where you store your 4-week rolling averages using AVERAGE formulas. Reference those averages in the OpenAI prompt as context: "Last 4-week average engagement rate was X%". This gives the AI summary week-over-week context without any additional API calls.

Next Steps

You now have a fully automated social media reporting workflow that pulls data, generates an AI-written narrative, visualises performance in real time, and delivers results to your team every Monday — with no manual effort after the initial setup.

From here, consider extending the workflow by adding Google Ads or Meta Ads spend data to correlate paid performance with organic reach. You can also replace the Slack delivery with an automated email using Make's Gmail module if your stakeholders prefer email over Slack.

If you would like help configuring this workflow for your specific platform mix, or want a more advanced version that includes AI-generated content recommendations alongside performance data, the team at Lenka Studio builds and maintains custom marketing automation systems for SMBs across Australia, Singapore, Canada, and the US. Reach out to discuss your reporting needs — we are happy to help you scope the right solution for your team's size and budget.