Keep your customers. Put AI into operations.
Articles from client work: how customers come back, how CX and CRM fit together, how to introduce AI and which rules it needs.
Contact pressure in CRM: why a cap per channel is too blunt
Many CRM setups limit contact with a single cap per channel. Five principles for contact rules that tell signals apart, plus a self-test for your team.
Dots or your own agent harness: which one fits you?
A hosted agent or a harness you build yourself? A comparison of runtime, memory, approvals and model choice, with a decision rule for mid-sized companies.
OpenAI Dots: what is actually new?
OpenAI Dots run continuously, with their own cloud computer and approval rules. What is new about them, where the limits are and what the situation looks like in the EU.
Proving the value of CX: why the business case needs a control group
A business case justifies the decision for a CX measure. Whether the measure works only becomes clear through ongoing measurement with a control group. Here is how to start with one signal.
The AI pilot was a success. Now what? What has to happen in operations after the rollout
The AI pilot is running. But who looks after the AI day to day, who notices errors, what does it cost to run and who is liable? Four questions and a handover checklist for decision-makers in mid-sized companies.
The containment rate as a CX KPI is a governance trap
Your chatbot reports a high share of handled conversations, yet customers are leaving. Why the containment rate misleads and what you should measure instead.
Your AI is only as good as your customer data: the 90-minute data check before your first tool
Why patchy customer data slows down AI projects in mid-sized companies, and how to check in 90 minutes whether your CRM is ready for its first AI tool.
Seen or assumed? The self-test that shows how much fiction is in your journey map
Mark every step of your process overview as "seen" or "assumed". That is how you find the transitions where you can lose customers without noticing.
AI documentation in mid-sized companies: three documents that let you answer questions at any time
Which AI is running in our company, who is responsible for it, and when did anyone last check? How mid-sized companies put their register, literacy records and customer touchpoints in order in four weeks and keep them up to date afterwards.
Measuring is not enough: why your health score programme needs an implementation mandate
Many customer programmes measure carefully and still change nothing, because nobody is allowed to fix the friction points they find. How to set authority and budget for implementation from the start.
Vibe coding in mid-sized companies: your company is already programming, and now you need a rule
Your business units are building their own tools with AI, usually without IT knowing. Why that is an opportunity, and why the four-eyes principle is the one rule you should introduce first.
Goodhart's law in customer metrics: why every activity figure needs an outcome beside it
A health score built from activity figures measures effort more than effect. Five examples of activity metrics, their missing counterpart, and a grid for checking your own reports.
AI in mid-sized companies: the numbers add up through augmentation rather than job cuts
Many AI business cases count on eliminated jobs and do not add up. Why augmenting small teams is the better calculation for mid-sized companies, with three cost levers and a guide for your own case.
Your CRM does not fail overnight: five test questions against its gradual loss of value
Workarounds, missing ownership, old processes in a new tool: how to tell whether your CRM still creates value. Five test questions for before and after the launch.
AI maturity in mid-sized companies: why many get stuck at level 2
Five levels of AI maturity in plain language. Why the jump from level 2 to 3 depends on culture, data access and backing from management, and how to tell that you have made it.
Journey management without proof of ROI: the problem lies in the data silos behind the dashboard
Almost a quarter of CX leaders struggle to demonstrate any financial return from journey management at all. Why this is a matter for management, and why a new dashboard changes little.
Data readiness before choosing a model: 8 questions that decide whether your AI project delivers
Whether your data is ready matters more than which model you pick. Eight questions for a first AI readiness check, doable in about an hour. For mid-sized companies without a BI team.
What Flight 401 teaches about customer early warning systems: six clear signals instead of colourful dashboards
A crew loses an intact aircraft because of a faulty light bulb. What that means for your customer dashboard, and which six early warning signals count.
Does your service chatbot say it is an AI? How to set up the notice in customer service
Text templates for chat, messenger and voicebot, a fixed owner and an approval rule: how a service team sets up and maintains the AI notice in customer contact, with a practical check in seven steps.
Why customer programmes fail on the data question, even when the concept is sound
Journey and loyalty programmes rarely fail because of the idea. They get stuck because nobody knows who is responsible for which customer data. Here is how to settle that on one page before the budget is approved.
AI in mid-sized companies: from testing to operations, and why the order of the hurdles matters
Many mid-sized companies have tested AI, fewer run it day to day. What Bitkom, DIHK and KfW show, and why competence is the first step.
Churn early warning for mid-sized companies: four signals you can spot in the data you already have
Churn often announces itself weeks in advance. How to read order rhythm, basket, service contacts and the contact clock from data you already have.
The 3-tier model for AI agents: which task belongs on which tier
Not every task needs the same AI effort. How to classify tasks: build it yourself, have professionals integrate it, or leave it with a person.
Reactivation Audit for dormant email flows in your CRM
Paused and forgotten email flows in your CRM are already paid for. How to inventory, assess and reactivate them in four steps.
Deployer check: why an honest AI inventory is the best entry point into AI transformation
Which AI is actually running in our company? How mid-sized companies use the deployer check to get an overview of their AI use, name owners and turn the result into rules for day-to-day operations.
Articles are written with AI support and editorially reviewed.