A 248% return is the kind of number that ends up on slides without its context. It comes from a Forrester Consulting Total Economic Impact study, commissioned by Microsoft and published in June 2024, that modelled what a composite organization could gain from deploying Power Automate over three years. It is a rigorous piece of work. It is also a model of one specific product in one specific kind of company, and Forrester says so plainly. Here is what it actually shows, and how to borrow its method.
How a TEI study works
Forrester interviewed seven representatives at six organizations already using the product, then built a composite organization from their profiles: 30,000 employees and about $10 billion in annual revenue. The financial model added up benefits and costs over three years, risk-adjusted the figures for uncertainty, and produced net present value and ROI. For the composite, benefits of $55.93 million against costs of $16.08 million gave an NPV of $39.85 million and the 248% headline.
Two things follow. First, this is a model of a very large enterprise, so the absolute numbers will not map onto a mid-size marketing team. Second, and more usefully, the ratios and the categories of benefit tell you where automation returns tend to come from.
Where the return came from
- Time released from repetitive, high-volume tasks. Employees involved in the highest-impact automation use cases saved roughly 200 hours a year, equivalent to a lift of around 10%.
- Further time savings from extending automation into simple internal apps and workflows, modelled at about 250 hours per affected user.
- A 20% reduction in the time required to build new workflows, because each automation made the next one cheaper.
Costs were not trivial: licenses, developer time, and training around 1,800 workflow builders plus a small team to run the platform. The return held up because the benefits compounded while the costs were largely fixed.
The caveat Forrester puts in bold
The study states that Forrester makes no assumptions about the ROI other organizations will receive, and advises readers to use their own estimates within the framework. That is the correct way to read any vendor-commissioned study, and it is the correct way to read this article: treat the number as an existence proof and the method as the takeaway.
What this means for creative and marketing teams
The tasks that drove Forrester’s return were document processing, data entry and routing. Marketing has its own version of that work around the work: rewriting briefs, resizing and versioning assets across channels, chasing approvals, checking claims against guidelines, and searching for the approved logo file for the fifth time. None of it is the idea. All of it takes time from the idea.
This is where an AI-native workflow earns its return. When brand knowledge lives in one structured place, drafts start from context instead of a blank prompt. When review happens next to the work, approval loops shrink. When claims and brand checks run before anything ships, legal review stops being a bottleneck. At Branday that structure is Vault, Spaces and Watchman, with our people setting direction and taking responsibility for the output.
How to evaluate automation ROI on your own team
- Baseline first. Measure hours per asset, cycle time and revision rounds for eight weeks before changing anything.
- Load the costs fully: tools, setup, training, and the review time your people still spend.
- Risk-adjust. Discount benefits you cannot yet prove; Forrester does exactly this.
- Look at payback period as well as ROI. A smaller return that pays back in months may beat a larger one that needs three years.
- Re-measure at a fixed checkpoint and compare against your baseline, not against the vendor’s composite.
A hypothetical illustration: a team spending 120 hours a month on adaptations, resizing and approval chasing that cuts this to 60 hours has released the equivalent of a third of a full-time role. Whether that is a good return depends on what the team does with the time and what the change cost. Those are the two numbers worth arguing about.
Want to see how the work around the work is handled in Spaces? Read about the platform or book a walkthrough.