IIOT: How to avoid Pilot Purgatory, lessons from the field
Keesjan (Case) Engelen
Chairman Titoma, Electr. Design & Mfg Colombia & China. On-Time & On-Budget since 2001
The problem McKinsey raised: "71% of industrial firms still stuck in Pilot Purgatory" is real, as confirmed by the hundreds of people reacting to my initial post.
@Abid Hussain nicely summed up how a Pilot From Hell typically goes down:
- Top management says: "We must get into IoT!" Budgets are approved and business units get money earmarked for pilots.
- Pilots are often designed around 'showcasing' the IoT provider's capabilities/distinguishing features, not necessarily solving the customers' problem. This may be due to the vendors' own agenda, or the fact the customer may not even have an identified problem - the biz unit was just told to do "an" IoT pilot.
- IoT program never gets rolled out because biz unit sees no value and is not willing to spend.
Technically it’s all doable, especially when using tried and tested hardware and platforms. The issues tend to be on the business case and the human side of things, and looking back many pilots should never have been started. With a very big thank you to all the professionals willing to share their experience, earned often the hard way, here's my summary:
Recommendations for Businesses
1. Solve a problem worth solving
- There is no point doing "something" in IOT just to show you can. Pilots which do not have a profitable business case just waste everybody's time, and will poison the well in a company for years.
- @Michael Riemer: Deliver an outcome not just information. Make sure the data is acted upon, leads to measurable benefits, increases company profitability and growth.
2. Keep it quick and simple
- Start with existing hardware and platform as much as possible. Remember this is just a Proof Of Concept (PoC), a quick win to build support and get the ball rolling. Once the business case is proven it is much easier to justify additional budget to further optimize the solution for a full roll-out.
- Use an experienced consultant. Your own manufacturing engineers are already pretty busy making sure the paying customers stay happy, and are unlikely to give priority to an internal project. It’s not easy to keep up with all the developments in this field, and solid implementation experience does help.
3. Manage the Human Factor
- Avoid using trendy buzzwords such as IOT, AI etc. Simply talk about the “Increase Up-time to 99.9%†project, and you sidestep a lot of resistance.
- Assure the support of the workers in the unit involved, what’s in it for them? Especially for a first project it makes sense to pick a problem everybody wants to see solved.
- Before kickoff make sure all stakeholders agree on how to measure success: have clear primary objectives and a strong verification strategy for each objective.
- Clearly define roles & responsibilities. Have a C-level exec put his name on the line. @Ahmed Sharaf: Without an executive evangelist you have no one to communicate the success of the project to the rest of the business. Put responsibility with the head of the unit concerned, not with IT or some "Chief Digitization Officer" who may be perceived as pursuing a hobby.
Recommendations for Solutions Providers
- Don’t do pilots for free, if the client is not willing to pay at least something then you have not found a problem they feel is worth solving, and you're unlikely to see money even if the pilot is successful
- Qualify the opportunity, learn to say no. @Samir Djendoubi: Do not launch a pilot if there is no top management support and budget for the real project. @Letha McLaren: The PoC should be simply a "gate" to moving forward with the rest of the business plan.
- Define success criteria upfront, set a time frame and get a volume commitment upon success.
- Consider focussing on smaller companies. @Nick Pummell: In our experience the SME companies are much more prepared to push ahead with disruptive technologies, are focused on growth and have a lot less decision makers to overcome. The bigger companies tend to be more difficult due to the sheer number of stakeholders - everything takes forever to get sign off - if ever.
- Focus on becoming an expert on a specific vertical, or on a specific building block. As a fresh start-up desperate to prove yourself you may have to accept to do a pilot for little money, but then at least make very sure the learnings will strengthen your expertise. @Philippe Lambinet: Reject pilots that are too specific to a customer or that are outside of your main market focus. Don’t spread yourself too thin.
PS Again many thanks to all contributors! I am not a pilot consultant, so I consider myself fairly objective. Further tips on how to get things moving quicker are very welcome, please comment below!
This article first appeared on the Titoma blog as IIOT: How to avoid Pilot Purgatory.
Case Engelen is CEO of Titoma Design Ltd. and has been designing and manufacturing electronics in Taiwan & China for 20 years. If you consider a custom product you should read China ODM factory designs: 7 pitfalls to avoid. If you like this article, well, please like it!
Experienced Embedded Systems Engineer
2 å¹´Great article. As Edwin Land said "My motto is very personal and may not fit anyone else or any other company. It is: Don't do anything that someone else can do. Don't undertake a project unless it is manifestly important and nearly impossible."
Head of Strategy. Create sustainable impact for the energy, industrial, and manufacturing sectors. I realize sustainability through profitability and add business value with APIs, IoT, real-time data, and software.
2 å¹´this is the most important one -> ?Solve a problem worth solving.... and more important; solve a problem that is worth solving and easy to explain. Pilots are often stuck in explaining how beautiful and complicated the technology is. A simple and understandable business case is often the most convincing.
AI | Solution Architect | Developer | Advisory Board Member
4 å¹´I really enjoyed this article. Not just for any IoT project but all products developed by a company. Solve the customers problem! Seems simple enough but truly curious why it is so hard to do.
CEO - Trackting
5 å¹´Great and full of meaning! I should have read it several years ago....most of the IoT today is made by small companies (for free) for big companies to show something connected at tradeshow....sad
Global Delivery Manager || Digital Cockpit || ADAS
6 å¹´Wow !! Very Well articulated ??