Ten logins to run one plant

What the customers were actually doing

The most telling findings were behavioural rather than stated. Customers were exporting data out of the supplier tools into spreadsheets, building charts by hand, and drawing their insight there. A maintenance executive described people who had stopped looking at alarms altogether because there were too many of them, and who wanted one root cause instead.

And the logins. Monitoring, motors and drives all carried separate dashboards, so operators signed in repeatedly through a working day. The question the customers asked back was the obvious one. If these are all your solutions, why can we not log in once.

None of that is a feature request. It is a set of workarounds, and workarounds are what a user builds when the product has stopped being worth arguing with.

The cost sits in training, and the customer pays it

One core platform was reported to take around fifteen days of training plus remote assistance before it was usable. The same platform is highly configurable, which customers named as both the strength and the reason the training is unavoidable, because almost nothing arrives preconfigured.

On another product, a customer service delivery team was reported to need over six months of learning time, with mistakes made on the job in between. Both figures come from interviewees rather than from measurement, and they are single source, so they are indicative rather than established.

Two products were described by users as visually obsolete, in language comparing them to desktop software from the mid nineties. A third was praised for its calculation engine and criticised for presenting the result badly, which is the clearest possible statement that the engineering was never the problem.

Four trends, five requirements

The interviews were reduced to four digitalisation trends, next generation analytics, a digital workforce, automation, and the convergence of information and operational technology. Each maps to a user experience requirement the supplier is not currently meeting: data visualisation, mobility, preconfiguration and customisability, cybersecurity, and a consistent experience across products.

The supporting evidence is third party and should be read that way. Process manufacturers were estimated to capture only 20 to 30 per cent of the potential value of data analytics where visualisation is absent. Connected workers were expected to contribute up to 85 million dollars of productivity gain in mining and metals. Automation was projected to remove up to 4 per cent of the total cost base in pulp and paper. All three are industry projections, none is a client measurement.

Cybersecurity turned out to be a purchasing criterion rather than a technical one. These customers treat operational knowledge as competitive advantage, so where the data sits decides whether they buy.

Why it had stayed this way

The root causes were internal and they were stated without softening. There were no user experience specialists at business unit level. User experience was not on the product development agenda. Design guidelines existed and were adopted voluntarily, which makes performance against them unmanageable.

One digital lead said communication was very poor and that they had never spoken to the design team. The consequence showed up as duplicated build: two business units had each produced a condition monitoring tool, and neither knew the other existed.

There was also a hard platform constraint. Several products are tightly coupled to a base automation system, so one of them cannot be redesigned on its own. Fixing the interface on a single product is not available as an option, which rules out the cheap answer.

The recommendation was not a redesign

A redesign would have been the expected deliverable and it would have been the wrong one, because the same interface would have come back on the next product. The recommendation was a four stage capability build instead.

First, get design into the product development process and make the case internally that it is a purchasing criterion for customers, not a finish. Second, build design capability inside each business unit and involve designers with customers early. Third, create the mechanism for units to share what they know, since they currently operate independently and have no incentive to. Fourth, put user experience into the company own stage gate model and track it with measures, so it is managed rather than encouraged.

That is a slower answer than a redesign and it is the only one that holds.

What this study can and cannot carry

Eleven interviews, of which two were customers. That is a thin customer base and it is stated here for the same reason it was stated in the deck. The five requirements are identified from those interviews, not validated by market research, and any claim built on them has to keep that word.

There is no measured result in this engagement. No product was redesigned, no capability was built and no improvement was recorded during the work. What it produced was a diagnosis specific enough to act on and a sequence that starts with the cheapest step.