Case Study
From 331 Hours to 15 Minutes
How a UK PR agency turned link building prospecting into a system that surfaces qualified prospects before the first morning coffee hits
Less than 1% of prospecting time was value-added
331 hrs
Cycle time per prospecting round
<1%
Value-added efficiency across the entire process
2.5 hrs
Actual value creation out of 331 hours
17+
Manual steps per prospect, from keyword research to send
PR prospecting is a time consuming task where each client needs their own keyword research, competitor analysis, backlink checks, Google Maps searches, email finding, draft writing, and outreach. 17+ manual steps per client. 331.75 hours of cycle time. And less than 1% of that time was actually creating value. The rest was clicking, copying, cross-referencing, and waiting.
Where the hours were hiding
| Step | Cycle time |
|---|---|
| Reciprocal opportunities | 2 min |
| Keyword research | 15 min |
| Competitor analysis | 20 min |
| Backlink analysis | 2 min |
| Search operators | 45 min |
| Google search across sites | 45 min |
| Compare previous work | 2 min |
| Google Maps research | 10 min |
| Find missing emails | 1 min |
| Draft communications | 1h 30m |
| Find relevant page to link | 5 min |
| Weekly competitor check | 10 min |
| Send list for approval | 15 min |
| Upload to outreach software | 10 min |
| Quality check each prospect | 5 min |
| Check each preview | 5 min |
| Send communication | 1 min |
Manually clicking on google pages and reporting on the validity of a page is extremely laborious and slow. A lot of that process can be done using a computer to automatically evaluate if a page is a good fit for a client. The checks, the clicking, the waiting... People deserve better systems to do their work.
The Lean AI Method
1. Discovered
Mapped every step of the prospecting process. Found 331.75 hours of cycle time with only 2.5 hours of value creation.
2. Designed
One system to replace 17+ manual steps: automated multi-source research and automatic page scoring for client fit.
3. Deployed
Built end to end. The clicking, cross-referencing, and waiting now run as a single automated pass.
4. Scaled
In production through year one. The prospecting round that once took 331 hours now runs in 15 minutes.
Why process comes first
We mapped the value stream before writing a line of automation. The map showed under 1% of the 331 hours was adding value — once you can see that, the build is the easy part.
What we built
Automated research
Multi-source data pulls replace 7 manual research steps
Prospect scoring
Automatic quality and client-fit checks eliminate manual vetting
Bulk page analysis
3,000+ pages evaluated for fit in a single 15-minute run
Before vs after
| Before | After | |
|---|---|---|
| Manual steps | 17+ | 0 |
| Cycle time | 331 hrs | 15 min |
| Efficiency | <1% | 100% |
15 minutes from data to prospect
£47k
Recovered in year one
99.9%
Reduction in cycle time
3,000+
Pages analysed for client fit in 15 minutes
0
Manual steps left in the prospecting pipeline
331 hours of manual work eliminated per cycle. Let people get on with the client work.
“You can't automate what you haven't mapped. The value stream map showed less than one percent of the work was adding value. Once you can see that, the build is the easy part.”
Sound familiar?
- Is your team running manual prospecting workflows with dozens of clicks per lead?
- Are you repeating the same research from scratch for every client?
- Do you know what percentage of your process is actually creating value?
- Have you tried automating parts of it but never mapped the whole thing end to end?