Automating business processes: where do you start when you don't know where you stand?
The tools are in place. The licences are paid for and the implementation is done. And yet leads keep falling through the cracks between systems, sales spends too much time on CRM admin, and marketing struggles to show which efforts actually contribute to pipeline or revenue. It’s a familiar reflex: a new tool, an extra integration, another dashboard. But that's rarely the answer.
With over 20 years of experience and working alongside more than 1,500 brands across Europe, we've come to one clear conclusion: most automation projects don't fail because of the technology. They fail because organisations start with the wrong question. Not "what's the problem?" but "which tool?" Processes get automated before they've been cleaned up. Or the whole thing gets built on data that simply isn't reliable.
Before you start automating business processes, you need to be honest about where you actually stand. Not in theory, but in reality. In this blog, we walk you through five strategic questions to ask before you start or expand your business automation efforts. And how the answers help you assess your organisation's automation maturity.
)
)
Five questions to assess your organisation's automation maturity
Question 1: Do you know what percentage of your CRM data is actually usable?
When we talk about usable data, we don't mean data that's been entered or is simply present in a system. We mean data that's current, complete, linked to the right records, and available for automation. Most organisations estimate their data quality at 70 to 80 per cent. An audit typically reveals it's closer to 30 to 40 per cent.
The gap comes down to definition. Data can exist without being usable. A contact record with a name and email address is present. A record with a valid email address, linked to the right company, with a job title filled in, a known source, and a recent interaction: that's actually usable. That distinction determines whether automation is even possible.
A real-world example: one organisation had 2.3 million contact records in their data lake. They were proud of that volume. After deduplication, cleansing, and validation, 340,000 usable profiles remained. The "rich customer data" turned out to be an illusion.
)
Question 2: Can sales, marketing, and service find the same customer in the same way?
Are sales, marketing, and service working in the same systems? Or does everyone have their own spreadsheet, their own naming conventions, a parallel system of their own? In many organisations, the honest answer is no. Sales exports a weekly overview from the CRM because the default view doesn't work for them. Marketing operates from a segmentation tool that's slightly out of sync. Service searches by ticket number because the customer link is unreliable.
This isn't a technical problem. It's a symptom of systems that don't fit the way people work. And ways of working that have therefore developed outside the systems.
It happens because adoption only enters the conversation once the implementation is done. That's too late. Adoption isn't a phase at the end of an implementation. It's a prerequisite at the start of a successful automation journey.
)
Question 3: When did you last delete a workflow?
Most organisations only ever add. Every campaign gets a new workflow, every edge case gets a new branch, every new team member builds their own variant. Until nobody knows what's running or why.
In a recent project, we found 47 active workflows in a marketing automation platform. It turned out that just 12 workflows were doing 80 per cent of the work: welcome sequences, abandoned cart flows, post-purchase flows. 23 workflows were duplicates or campaign-specific variants that had never been cleaned up. 12 were technically active but running on segments with no contacts, doing nothing useful. After the clean-up, 15 workflows remained. The system ran faster, the team understood what was happening, email deliverability improved because fewer unnecessary messages were being sent, and there was finally space to build new, well-considered workflows.
The point: automation that isn't maintained becomes automation that gets in the way. Regular clean-ups aren't a luxury, they're a necessity.
)
Question 4: How many hours a week do your people spend serving systems instead of doing their actual work?
Entering data, building reports, searching for information that should be somewhere but isn't. In many organisations, a significant chunk of working gets eaten up by searching, checking, and manually processing information. In fact, more than 40 per cent of employees spend over ten hours a week on repetitive tasks like admin and manual data processing.
Automation is supposed to reduce that friction. The reality is that poorly implemented automation sometimes increases it, because people end up wrestling with systems that aren't intuitive or don't fit their day-to-day work.
Business automation isn't an IT project. It's an organisational question. Who does what, and why? Who owns a process once the implementation is done? Without answers to those questions, you end up with a system that works technically but fails to take hold across the organisation. Approach automation as an organisational question, and sales gets more time to sell and marketing gets more space to optimise campaigns.
)
Question 5: If you switched platforms tomorrow, how much of your current setup would you rebuild exactly as it is?
A platform migration is the acid test. What would you rebuild? And what would you leave behind?
The question reveals which parts of your current setup genuinely add value and which have simply grown over time without any deliberate choice. It's a reality check for anyone who thinks the current setup ‘just works’.
Processes in organisations are rarely designed with intention. They simply evolve over time. Someone started doing something a certain way, others followed, and after a while nobody remembers why it works that way. Those processes, with all their inefficiencies, workarounds, and exceptions, then get automated. Which creates more problems than it solves.
"You can't build on mud. Everyone says they're data-driven until you ask them to show you the data. Then it turns out there's a layer of gold on top and mud underneath. You can't build on mud."
Steven Van Duyse, HubSpot Domain Lead
Automating business processes: where do you stand?
If you've answered those five questions honestly, you already have a clearer picture of where your organisation stands. To go further, it helps to think in terms of maturity levels. The automation maturity model gives you a framework for positioning where business automation sits in your organisation today.
)
Level 1: Ad hoc No structured CRM or automation process. Standalone tools, manual processes, teams working in silos. Data is scattered across the organisation and there's no clear view of what's working and what isn't. The defining characteristic of this level isn't the absence of tools, it's the absence of structure.
Level 2: Basic The CRM is used for transactional activities but in reality acts as little more than an expensive address book. System adoption is low. There's little to no automation and no clear picture of the end customer.
Level 3: Structured Workflows and automation are properly set up and there's a 360-degree account overview. Processes are standardised and data governance is defined. The challenge here shifts to adoption and maintenance.
Level 4: Optimised Business automation is fully integrated across functions and teams. AI-driven lead scoring and cross-functional workflows are in use. System adoption is high and continuous improvement is the norm.
Level 5: Predictive The CRM operates as a strategic growth engine. Predictive AI and hyper-personalisation are standard. A typical characteristic of this level: a team of eight performs like a team of sixteen, thanks to smart use of automation and AI.
)
Where are you at?
A quick way to gauge your organisation's level:
Do you go beyond first-name personalisation? → Level 1+
Do you have segment-based rules that drive behaviour? → Level 2+
Are profiles updated in real time based on behaviour? → Level 3+
Is the experience consistent across email, website, ads, and service? → Level 4+
Is every interaction unique per person, cross-channel, in real time? → Level 5
Most organisations we speak with are somewhere between levels 2 and 3. The tooling is there, but they're struggling with adoption, data quality, and cross-functional alignment. The gap between level 2 and level 5 isn't closed by better technology. It takes better architecture, organisation, and discipline.
)
Automating business processes starts before the tool
If you've answered the five strategic questions in this blog honestly, you now have a better sense of where you stand. And knowing that, you probably want to know how to take the next step.
The good news: you don't have to solve everything at once. Start with the foundation: People, Process, Information, it’s what we call the PPIT framework. The 'T' for Technology comes last, deliberately. Bad processes don't get better through automation. They get worse, faster. And if your data isn't right, you're automating the wrong decisions.
In the blog 'The PPIT framework: the four factors that determine whether business automation succeeds or fails', we go deeper into the PPIT framework: how to get the right sequence and avoid automating chaos.
Whitepaper: The four factors that determine whether business automation succeeds or fails
Download the whitepaper CTRL+SHIFT: The four factors that determine whether business automation succeeds or fails and discover how to align People, Process, Information, and Technology - so automation works not just in a demo, but in the real world, day in, day out.
)