PCE, a Production Chain Erosion, is a chronic state where the company's production chain runs sub-optimally and is prone to a delivery outage. Small companies are not infected, but as the size scales up, the risk multiplies. PCE can often be traced back to a prolonged shortage of essential software features and a lack of interest on data hygiene. There's no vaccination, but taking in a designer to your software project, well before buying or coding business software, pays back quickly.
Before diving too deep into the solutions, let's rewind. What is a business software? When classifying software, I often use a commonplace trichotomy. Most of the software we use can be called consumer software: a public transport app on your mobile or a weather forecast in your browser. Often these are single-purpose apps with only a few features. Then there are personal productivity tools like word processing or photo editing: rich on features, while still based on one user role. The third category is business software: stuff that organizations use to handle their resources – labor, capital or land, as in a classical economy. You name it: Enterprise Resource Planning, Product Lifecycle Management, Customer Relationship Management, Workforce Management, Patient Information system, time tracking, payroll, billing, invoicing or bookkeeping. Apps that many of you use at work, and which too often excruciatingly slow you down.
Based on my personal experience, people buying business software are often not aware of how complicated the requirements of features are in the end.
I recently had a thought-raising discussion with my father, who proved to be one of the first guys visioning a paperless office around the globe. Not that long ago in the 1970's they wrote their own databases and built drivers from scratch to get those bytes printed when paper-oriented people insisted. He did software when there was no such thing as a display, let alone a graphical user interface.
This discussion led me to a chain of thoughts.
Consumer apps of the day feel easy and even delightful to use for numerous reasons. Huge markets and serious competition drive companies to invest in concept design and user experience. There are often just a few features, which makes design work easier and learning curve short. Simple software can be designed for one user role. Consumers make the decision to buy, stay or leave themselves – not a board.
In business, the core software is heavy on data and features. Users come in numerous roles. Decisions to buy or back off are often made by a board whose members don't always use the products themselves. Once a core business app is in production, changing it on the fly can be extremely difficult, slow and expensive.
Data and the things you do with it (features) were there already before the graphical user interface.
With consumer apps, the data- and the feature-related decisions are usually taken care of by the businesspeople. Designers help with the feature's path details and the general look and feel. This is logical, as easy-flowing paths and look and feel make the difference on the market.
With business software, look and feel has a minor role, and work usually gets done even if the paths do not flow smoothly. Having the right features and the ability to handle incoming and outgoing data makes the difference. Based on my personal experience, people buying business software are often not aware of how complicated the requirements of features are in the end.
Against this background, I present two preventions against PCE, starting from the most important one.
Before doing market research or benchmarking, map the features that are needed while executing that exact part of your production chain.
Prevention First aid: Map your Features
When an organisation needs new software, it has three ways to go:
1) Deploy the features from a deprecated ecosystem that's probably already in use.
2) Believe in a newcomer, be it an ecosystem, a bunch of single-feature software or a combination.
3) Build your own.
Sometimes the deprecated is good enough. Cases 2 and 3, however, will eventually lead to a software project –buying, building or a combination. The usual procedure has predictable steps. Someone first gets a nakki (Finnish translation for an unpleasant job) of executing a service provider survey and presenting a few of the best options. One option is picked based on 2-4 factors: license fees, fit on existing tech, sometimes information security/privacy, and a belief that the app does the job.
Fees are fairly easy to calculate, even with growth in mind. For a tech fit, a software architect may carry out technical due diligence. For information security, there are specialists as well.
The most camouflaged business risks rely on the belief that the app does the job.
Before doing market research or benchmarking, map the features that are needed while executing that exact part of your production chain. If you're uncertain how your production chain builds up, first carry out a service blueprint or other document that shows the actions needed to make a delivery happen.
For example, if you need a tool to handle working hour entries from multiple subsidiaries, walk through the whole process phase by phase with the people currently handling the entries in their ways. Name every step, and preferably note the data sources and format. Generalize the steps so that they're not bound to current tools but to the user's needs (in software development these are called user stories and epics that bundle stories). Transcribe a memo.
When you've done this, you're able to compare service providers based on the real need, not a belief. Into the bargain, you've got a work introduction/training document. Furthermore, if you find out there are no apps for the need in the market, you've already done a big part of the software specification.
To map features an easy and professional way, take a designer to do it for you. Besides excellent visualists, there are dedicated user research- and concept designers. Figuring out rough user stories is their daily bread, and some wizards master advanced methods like goal-directed design simulation, in case a process must be scavenged in-depth.
Modes of action
Significant increment of productivity: employees don't have to patch their workflow.
Operational reliability: should an employee vanish, successor doesn’t have to spend the first 2 months building patches from scratch.
Unexpected benefits: a roster that used to be an excel-print on a wall, is suddenly online and up-to-date. Shiftwishes can be proposed within the app instead of emails or whatsapp groups.
Significant savings on software development. (This topic is worthy of another post).
Download our Design Sprint e-book as a pdf here.