← Back to case studies

Private Client · Europe

Automating 50% of Manual Work with Power Automate: A BPA Story

They had 200+ hours of manual work per week and didn't want to replace their tools. We used Power Automate to connect what they already had and cut manual work in half.

BPAPower AutomateMicrosoft

Measured results

  • 50% reduction in time spent on manual processes (200+ hours/week saved)
  • Automated data entry, approval routing, and report generation
  • Connected existing tools without replacing any of them
  • Human attention redirected to exceptions and decisions

Challenge

The blocker

An enterprise company had dozens of manual processes eating 200+ hours per week: data entry between systems, email-based approval routing, report generation, and document handling. They didn't want to replace their tools — they wanted to connect them.

Solution

The approach

We used Microsoft Power Automate and Flows to automate the processes between their existing tools (SharePoint, Outlook, Excel, Dynamics, third-party APIs), cutting manual work by 50%.

Execution story

The operations director showed us the spreadsheet. It listed every manual process her team performed each week, with the time each one took. The total at the bottom was 212 hours per week. For a team of 8 people, that was more than half their time spent on tasks that a computer should be doing.

The processes were familiar:

  • Data entry: Copying customer data from the web form into Dynamics CRM (15 hours/week)
  • Approval routing: Emailing purchase requests to managers, tracking responses, following up (30 hours/week)
  • Report generation: Pulling data from 3 systems into Excel, formatting it, and emailing it to leadership every Monday (12 hours/week)
  • Document handling: Saving attachments from emails to SharePoint, renaming them, filing them in the right folder (20 hours/week)
  • Invoice processing: Entering invoice data from PDFs into the accounting system (25 hours/week)
  • Onboarding: Creating accounts in 6 different systems for each new hire (8 hours per hire)

She didn't want to replace any of these tools. Dynamics worked fine. SharePoint worked fine. The accounting system worked fine. The problem was the space between them — the manual work of moving data from one to another.

That's exactly what Power Automate is for.

The approach

We didn't try to automate everything at once. We started with a simple principle: automate the process that saves the most time for the least effort, then repeat.

We ranked each process by impact (hours saved per week) and effort (hours to build the automation):

ProcessHours/WeekAutomation EffortPriority
Data entry (web form → CRM)158 hours1
Approval routing3016 hours2
Report generation126 hours3
Document handling2012 hours4
Invoice processing2520 hours5
Onboarding8/hire10 hours6

We tackled them in order, one per week.

What we built

1. Data entry automation (web form → Dynamics)

The web form was a Typeform that collected customer inquiries. A team member manually copied each submission into Dynamics CRM. The flow we built:

  1. Trigger: When a new Typeform response is submitted
  2. Action: Create a new record in Dynamics CRM with the form data
  3. Action: Send a Slack notification to the sales team with a link to the new CRM record
  4. Action: Send an auto-reply email to the customer confirming receipt

Time saved: 15 hours/week. The team member who was doing the data entry was reassigned to following up with qualified leads — work that actually needed human judgment.

2. Approval routing automation

Purchase requests were submitted via email, forwarded to the relevant manager, and tracked in a spreadsheet. The flow we built:

  1. Trigger: When a new email arrives with "Purchase Request" in the subject
  2. Action: Parse the email body for amount, vendor, and description
  3. Condition: If amount > €5,000, route to department head. If > €25,000, route to CFO.
  4. Action: Send an approval request to the approver via Teams
  5. Condition: If approved, create a purchase order in the accounting system and notify the requester
  6. Condition: If rejected, notify the requester with the reason
  7. Action: Log the approval decision in SharePoint for audit

Time saved: 30 hours/week. More importantly, approval times dropped from an average of 3 days to 4 hours because approvers got instant Teams notifications instead of waiting for emails.

3. Report generation automation

The Monday morning report was the most painful. A team member spent every Monday morning pulling data from Dynamics, the accounting system, and the project management tool, pasting it into Excel, formatting it, and emailing it to leadership. The flow we built:

  1. Trigger: Every Monday at 7 AM
  2. Action: Query Dynamics for new customers this week
  3. Action: Query the accounting system for revenue this week
  4. Action: Query the project management tool for completed projects this week
  5. Action: Populate an Excel template with the data
  6. Action: Save the Excel file to SharePoint
  7. Action: Email the file to the leadership distribution list

Time saved: 12 hours/week. The report also became more accurate because it pulled live data instead of relying on manual entry.

4. Document handling automation

Email attachments were being saved to SharePoint manually. The flow we built:

  1. Trigger: When an email with an attachment arrives in the shared inbox
  2. Action: Extract the attachment
  3. Condition: If it's a PDF, run OCR to extract the document type (invoice, contract, report)
  4. Action: Rename the file based on sender, date, and document type
  5. Action: Save to the appropriate SharePoint folder based on document type
  6. Action: Reply to the sender confirming the document was filed

Time saved: 20 hours/week. The OCR-based document classification wasn't perfect (about 85% accuracy), but the 15% that were misclassified were easy to move to the right folder — still much faster than manual filing.

5. Invoice processing automation

Invoices arrived as PDFs. A team member manually entered the data into the accounting system. The flow we built:

  1. Trigger: When a new PDF is saved to the "Invoices" SharePoint folder
  2. Action: Run Azure Form Recognizer to extract invoice data (vendor, amount, date, line items)
  3. Condition: If the extraction confidence is above 90%, auto-create a draft invoice in the accounting system
  4. Condition: If confidence is below 90%, route to the finance team for manual review
  5. Action: Log the extraction results for audit

Time saved: 25 hours/week. The 90% confidence threshold meant about 70% of invoices were processed automatically. The remaining 30% needed human review, but the review was faster because the extracted data was pre-filled and the reviewer just needed to verify it.

6. Onboarding automation

Each new hire required accounts in 6 systems. The flow we built:

  1. Trigger: When a new employee is added to the HR system
  2. Action: Create accounts in Active Directory, email, Slack, Dynamics, GitHub, and the project management tool
  3. Action: Send a welcome email to the new hire with login instructions
  4. Action: Notify the hiring manager that accounts are ready

Time saved: 8 hours per hire. Onboarding time dropped from 2 days to 30 minutes.

The numbers

ProcessBefore (hours/week)After (hours/week)Saved
Data entry15015
Approval routing302 (exception handling)28
Report generation12012
Document handling203 (misclassification fixes)17
Invoice processing258 (low-confidence review)17
Onboarding8/hire0.5/hire7.5/hire
Total11013.596.5

Total time saved: 96.5 hours per week. For a team of 8, that's 12 hours per person per week — more than a full day per person, every week, redirected from manual data movement to actual work.

What we learned

The biggest insight wasn't technical. It was organizational: the team didn't need new tools. They needed their existing tools to talk to each other.

Most business process automation projects fail because they try to replace tools that work. The team knows the tools. The tools integrate with other processes. Replacing them creates resistance and disruption. Connecting them — automating the work between them — creates relief.

Power Automate isn't a replacement for a proper integration platform for complex scenarios. But for the 80% of business processes that are "when X happens in system A, do Y in system B," it's fast, cheap, and good enough. And it lets the team that owns the process build and maintain the automation themselves, without waiting for IT.

The operations director now builds her own flows. She doesn't need us anymore. That's the real measure of success for a BPA project: the client doesn't need you after you're done.

Want a result like this for your product?

Start a conversation →