door Diederik Martens |
Last Updated: 19 februari 2026 Your Automation Roadmap Starts With How Work Really Happens
The Moment I Realized “AI-First” Was the Wrong Starting Point
I just returned from a conference and you can’t escape the fact that everything is about “businesses want to go AI-first,”. I feel both excited and uneasy. Excited, because the ambition is real. Uneasy, because I have already seen many teams try this and fail. Not because AI doesn’t work. But because AI was being asked to fix something it didn’t understand: the real way work happens inside the organization.
It’s easy to assume AI will clean up a messy process for you. But once you dig into the details, you often realize the process you were hoping to automate probably shouldn’t even exist in the first place. Or it exists, but not the way you think. Or it exists differently in the minds of the people doing the work versus the people designing it.
What I learned is simple: before any team can successfully adopt AI or agents at scale, they need something far more fundamental. A truthful understanding of how work truly flows today. The real work. The invisible work. The “we’ve always done it this way” work. And that realization is what sparked a completely different way of building an automation roadmap.
Why Teams Jump to AI (And Why It Fails Without Process Reality)
Most teams don’t start with automation because they love complexity. They start because something hurts. A repetitive task. A slow handover. A manual process that was only meant to be temporary. And when AI enters the picture, it becomes incredibly tempting to skip ahead. AI looks like a shortcut. A chance to bypass fixing the underlying process. But here is the uncomfortable truth I’ve learned repeatedly: you shouldn’t automate what shouldn’t exist in the first place.
I’ve seen teams spend weeks automating steps that, when reviewed honestly, provided no value.
They were the results of old decisions, outdated approvals, legacy workarounds, or processes
designed around exceptions rather than the rule. AI, in these cases, doesn’t solve the problem. It accelerates the dysfunction.
And this is why so many AI attempts stall. AI can absolutely add value, but only if the basics are
solid. If you don’t understand how the work actually happens day-to-day, across marketing,
sales, and everything around them, you end up automating guesses instead of reality. This is what led me to develop my methodology called Chaploop.
What I Learned When I Started Checking How Work Really Happens
When I first began applying the Chaploop, I expected to uncover a few inefficiencies or minor
inconsistencies. Instead, what I found was something far more revealing.
Work, in reality, rarely resembles the diagrams, the SOPs, or the flowcharts created during
workshops. People don’t follow the ideal path. They follow the path that gets the job done. The
path they found yesterday that worked just well enough to repeat today.
I learned that a real automation roadmap starts with watching people work, asking honest questions, and noticing all the small frictions they’ve learned to live with. What surprised me wasn’t how many issues I found, but how consistently teams had been working around them without ever calling them out. Individuals become the glue holding broken processes together. They develop routines, hacks, and habits that are invisible in documentation but essential in practice.
This is where automation should begin. Not by perfecting the process on paper, but by understanding the process as lived by the people who rely on it every day.
Unexpected Patterns You Discover When Observing Real Work
Once you start observing real work with curiosity instead of assumptions, patterns emerge quickly. You notice how frequently people repeat tasks because “it’s faster to just do it manually.” You see how often information is retyped, reformatted, or hunted down across different systems. You discover that personalization is limited not by strategy, but by the lack of enriched data. You realize that the same campaign is rebuilt from scratch dozens of times because that’s simply how it has always been done.
I’ve watched marketers spend entire afternoons cloning programs and fixing assets manually. Sales reps often tell me they can’t find the notes from a call they had last week because those notes never made it into the CRM. I’ve seen targets lists being inserted from memory instead, leading to inconsistent data.
These patterns aren’t signs of failure. They are signs of reality. Reality is the best input you’ll ever get when shaping an automation roadmap. All those repeated tasks, manual searches, small errors, and poor handovers are signals. They show exactly where thoughtful automation, instead of AI for the sake of AI, can make work easier and more consistent.
Turning Reality Into Opportunities Through Use Cases
Once you uncover how work really happens, the next step is simply to capture where things
break down. You don’t start with tools or solutions. You start with observations. Every repeated task, every manual handover, every “I hope I didn’t forget anything” moment is a use case. It doesn’t need to be polished or technical. It’s just a step in the process that could work better.
When collecting use cases, avoid judging them. Don’t ask yet whether AI can solve it or whether automation is appropriate. The goal is to make the invisible visible. Common examples include unlogged meeting notes, campaigns rebuilt manually, gut based target lists, and social posts created from scratch. These aren’t failures. They’re signals of where improvement is possible.
Quantity over perfection. Capture it all first. Prioritization comes next.
How to Prioritize Use Cases (The Three Lenses I Actually Use)
Once you have your collection of rough use cases, it’s time to bring clarity and focus. Not everything should be automated. Not everything delivers equal value. And not everything deserves to be first.
After working with many scattered commercial teams, I’ve narrowed prioritization down to three lenses that consistently produce strong decisions.
Lens 1: Frequency × Time × Cost
This is the simplest and most objective way to spot high value opportunities. Ask three
questions:
- How often does something occur?
- How long does each time take?
- What is the cost of that time?
On paper, some tasks look tiny. But once you see how often they happen each month, you realize they take up far more time and energy than anyone expected. Take meeting summaries as an example. Writing up meeting notes and adding them to the CRM doesn’t seem like much. Maybe ten minutes. But across dozens of calls each week, for every person in sales, those minutes quietly turn into hours of lost time.
Lens 2: Human Error Reduction
Not every task is time heavy. Some are risk heavy. You can have steps that take only a minute or
two, but mistakes can create significant damage downstream.
Examples I’ve seen repeatedly:
- Incorrect contact enrichment leading to wrong segmentation.
- Inconsistent ICP tags resulting in poor targeting.
- Campaign configuration errors leading to mismatched audiences or broken links.
When you look at a use case through the lens of human error, you ask: - How often do mistakes happen?
- What does each mistake cost?
- How to recover from such a mistake?
Sure, automation saves time. But it also keeps things consistent. It reduces avoidable mistakes.
So, it also gives customers a better experience.
Lens 3: Business Effectiveness (Light but Important)
Some tasks don’t happen often. Some aren’t error-prone. But they directly influence revenue, customer experience, or strategic outcomes. This is where a simple impact-effort-matrix helps:
- Does this step influence lead quality?
- Does it improve personalization or relevance?
- Does it accelerate the buyer journey?
- Does it strengthen sales effectiveness?
Even if these use cases don’t top the efficiency chart, they often deserve a place in the roadmap
because they increase the effectiveness of teams. For most teams, efficiency and reducing errors highlight the biggest wins. But using all three lenses together gives you a simple, practical way to score and prioritize every opportunity you’ve identified.
Turning Priorities Into an Automation Roadmap
When your priorities are clear, you can finally start putting your automation roadmap together.
This is when the ideas stop feeling scattered and start forming something useful. And just to be clear, a roadmap isn’t a checklist of automations. It’s really about improving how the work flows in real life.
Start by grouping use cases into themes. For marketing and sales teams this could be:
- Data readiness: enrichment, tagging, classification.
- Campaign operations: building, launching, QA.
- Sales effectiveness: meeting notes, follow-ups, CRM hygiene.
- Segmentation & targeting: ICP matching, account curation.
- Content operations: planning, versioning, publishing.
These themes often reveal dependencies you wouldn’t spot otherwise. For example:
- You can’t automate advanced personalization until you have better data enrichment in
place. - Campaign auto-generation becomes far easier once naming conventions and briefs are
standardized. - Account prioritization improves dramatically once ICP tags and firmographic signals are
consistent.
This is where your roadmap starts to take shape. You place foundational improvements first. Steps that unlock capabilities further down the chain. Higher-value automations follow once the groundwork is stable.
The roadmap becomes a sequence of sensible steps:
- Fix or streamline the underlying process.
- Implement automations that remove repetition or reduce error.
- Add AI or agent-like components where human judgment is limited.
- Connect improvements so they reinforce each other.
One example from my own work illustrates this well.
A team wanted fully automated campaign builds and reviews. But when we checked reality, we discovered low quality campaignstandards, unclear campaign briefings, and inconsistencies in how campaigns were built by different people. The solution wasn’t to start with campaign automation. It was to start with standardization of example campaigns and campaign briefings. Once those were in place, the move to automated campaign builds became more natural.
Why Cycles Create Real, Lasting Improvement
Once your roadmap is clear, the next instinct is often to start a large project. But large projects
introduce risk, delay feedback, and make it harder for teams to adapt along the way. Real improvement comes from cycles, not big-bang launches.
Cycles allow you to deliver value quickly, learn from real usage, and adjust before investing more. Each iteration becomes a safe test: does this step actually help the people doing the work? Does it reduce effort? Does it improve accuracy? Does it remove friction?
This is why the Chaploop emphasizes short loops of checking, aligning, and performing. Each cycle is small enough to be safe, but impactful enough to build momentum. Teams stay in control. AI or automations take on predictable execution while people guide judgment and quality.
And the most powerful part? Cycles compound!
The first improvement often feels small. But each cycle reduces complexity, removes errors, and frees up time. And that new capacity becomes fuel for the next cycle. E.g. Once meeting notes are logged consistently, sales follow-up becomes more reliable.
Each improvement strengthens the next. And this is where the real productivity gains come from. Not from individual or personal automation, and not from personal AI use. It’s the compounding of many small, connected improvements that actually scales.
Final Thoughts: Start Your Automation Roadmap With How Work Really Happens
The pressure to “go AI-first” is real, and it will only increase. But the teams that benefit most from AI are not the ones who start with technology and company custom GPT’s. They’re the ones who start with reality.
They understand how work truly happens. They identify the friction. They prioritize what matters.
They build roadmaps grounded in truth, not assumptions. And they improve through steady cycles, not overwhelming projects.
AI and agents can dramatically accelerate work, but only when wrapped around clean, consistent, well-understood processes. The pathway to meaningful automation and scalable AI impact always begins with one simple question: How does the work really happen today? If you start there, your automation roadmap won’t just look good on paper. It will work in practice and it will keep getting better with every cycle.
And that is exactly why the Chaploop exists. It helps teams actually see how the work gets done. Not how they think it gets done. From there they can make small fixes. Then try things out and see what works. Then adjust along the way. After a while this just becomes the normal way of improving things and the improvements start piling up.
Graphics source: AI generated & Chapmanbright
We respecteren uw privacy, u kunt zich op ieder moment afmelden voor deze nieuwsbrief.