The DACH logistics market in 2026 is worth roughly 350 billion euros. The digitalization rate sits between 28 and 35 percent. Two-thirds of all logistics companies still rely on manual processes in key areas — phone calls for status updates, Excel spreadsheets for route planning, printed delivery notes that get manually entered into the system each evening.
The reasons are almost always the same. It’s not the technology. The technology exists, is affordable, and proven. It’s the execution: lack of internal capacity, fear of change, overwhelm from the sheer number of options, and the lingering memory of failed projects in the past.
At Exasync, we’ve completed two logistics projects in recent months, both productive in under four weeks. The philosophy: results over perfection, automation over software purchases, a working prototype over an 80-page requirements document.
A freight company with 40 vehicles had a specific problem: nobody reliably knew when the trucks would arrive. Fleetboard telematics was installed — so the data existed. But nobody sat in front of the web interface all day. Dispatchers worked by phone: call the driver, ask for their position, estimate arrival time, inform the customer. With 40 vehicles, that meant four hours daily just for status inquiries.
The approach:
What actually happened: The first version didn’t work. Fleetboard dynamically renders data — the web interface shows placeholders on initial load, and the real values only appear after one to two seconds. Power Automate Desktop was reading the placeholders instead of the actual data. The solution: explicit waiting for DOM rendering with dynamic wait times.
Second stumbling block: Fleetboard provides relative arrival times (in 2 hours 30 minutes) instead of absolute timestamps. Combined with time zones and daylight saving time changes, this single detail alone cost two days of development.
After three weeks, the system ran stable. Time spent on arrival communication: from four hours daily to zero. Dispatchers now use the freed-up time for route optimization — an area that had been chronically neglected because there simply wasn’t enough time.
A trading company receives 50 to 80 orders daily as PDF attachments via email. Two employees spent several hours each day manually typing order data into the ERP system. Error rate: an estimated three to five percent — wrong article numbers, swapped quantities, missed line items. Every error leads to incorrect deliveries, returns, and customer dissatisfaction.
The tech stack:
The devils in the details: Every supplier has a different PDF layout. Some use tables, some continuous text, some a mix. The solution isn’t one parser that handles everything, but a multi-stage chain that escalates. The system doesn’t need to handle 100 percent. 85 to 90 percent is enough — the rest goes to a human operator. Actual recognition rate after six weeks of optimization: 92 percent.
The result: instead of two employees each spending three to four hours daily on manual data entry, one person now spends 15 minutes each morning reviewing exceptions. The error rate dropped from three to five percent to under one percent — because automated validation catches errors that humans overlook.
This question comes up in every initial consultation. The honest answer: it depends. But the numbers help with orientation.
| Criterion | Enterprise WMS/TMS | Custom Automation |
|---|---|---|
| Initial costs | 50,000–500,000 EUR | 5,000–30,000 EUR |
| Ongoing costs | 2,000–15,000 EUR/month | 200–800 EUR/month |
| Time-to-value | 6–18 months | 2–6 weeks |
| Customizability | Limited (modules, configuration) | Complete (own code) |
| Vendor lock-in | High | Low (open-source basis) |
| Scalability | Excellent (above a certain size) | Good to very good (with architectural discipline) |
Rules of thumb:
At Exasync, we frequently see a combination: the client already has an existing system (often an affordable or self-built WMS), and custom automation fills the gaps. The PDF order processing above is a classic example — no WMS in the world has a parser for the individual PDF layouts of every supplier a mid-market trader works with.
1. Thinking too big: Trying to digitalize everything at once leads to analysis paralysis. One logistics company wanted to simultaneously digitalize route planning, warehouse management, invoicing, and customer communication. After nine months of planning, nothing was implemented. Better: identify one pain point, automate it, learn from it, then move to the next.
2. Buying software instead of understanding the process: A WMS costing 200,000 euros was implemented, but employees continue using Excel — because the WMS doesn’t reflect the actual process. Better: first document the current process, find bottlenecks, then decide whether software or automation is the solution.
3. IT department as the bottleneck: One to three IT staff are expected to handle daily operations AND digitalization. That doesn’t work. Digitalization requires dedicated resources or an external partner who doesn’t burden the operational IT team.
4. No monitoring: System goes live, nobody watches. Errors go unnoticed for days or weeks. Trust in automation erodes. Better: dashboard with KPIs from day one. Heartbeat checks, automatic restart on failure, error rate tracking. At Exasync, we use Supabase Edge Functions for this, checking system health every five minutes.
5. Ignoring change management: The technology works, but employees don’t use it. Because they weren’t involved, because they’re afraid, because they don’t see the benefit. Better: involve employees early, create quick wins, communicate transparently what the automation does and doesn’t do.
Both projects — fleet automation and order processing — share a monitoring infrastructure. That’s not coincidence, it’s principle: monitoring isn’t a feature you bolt on at the end. It’s a fundamental prerequisite that runs from hour one.
The architecture:
Exasync’s B-Drone approach goes even further: a dedicated mini PC (Dell OptiPlex Micro) runs 24/7 and executes automations that can’t or shouldn’t run on a client’s server. Cost: a one-time 400 euros for hardware, 15 euros per month for electricity and internet. More reliable than any cloud VM and cheaper than any managed service.
No theory — this is the roadmap we actually follow with clients:
Weeks 1–2: Identify the problem. Ask employees: Where do you lose the most time? Not management, not the IT department — the people on the ground. Typical candidates in logistics: manual data entry, phone-based status inquiries, customer communication during delays, invoice reconciliation, route planning.
Week 3: Assess feasibility. Good candidates for automation are processes that meet three criteria: repetitive (at least daily), rule-based (clear if-then logic), and with digital input data (email, web interface, database).
Weeks 4–6: Build a prototype. A working prototype beats any concept paper. The prototype doesn’t need to be perfect. It needs to demonstrate that the automation fundamentally works and delivers measurable value.
Weeks 7–10: Pilot operation and optimization. Parallel operation of old and new processes. Collect feedback, handle edge cases, improve recognition rates. This is where the bulk of optimization happens.
From week 11: Production. Old process retired, monitoring active, escalation rules defined. The first measurable ROI is in — and with it, the argument for the next project.
Exasync itself is the best proof of this approach: one founder, 50 AI agents, bootstrapped, Estonian company, 10,000 euros in revenue in three months. Not despite the iterative approach, but because of it.
Free initial consultation — an honest assessment in 30 minutes, no sales pitch. Read more: 5 Processes Every SME Can Automate Right Away | AI Automation Agency | Industry Solutions.
