We are regularly in contact with companies that complain their new product is in ‘product development limbo’… it was planned to take 3 or 6 months, but it turns out that after 9 to 12 months there are still gnarly technical challenges to surmount.

Here is a typical scenario that seems to play out quite often:

• Startup company forms to bring a new product to market, has an idea, and does some preliminary research & analysis;
• Startup company can’t do some/all of the engineering design work, so they look at the alternatives;
• They feel the product will be made in China, and they like the idea of having the manufacturer handle the product development and then production – the idea that ‘the same engineers take care of it all’ is appealing;
• Some of the manufacturers approached see the project as appealing – maybe they already have experience in the same technologies and they think it will be easy, maybe they see that product space in general as interesting, maybe they see the startup is well funded and promising;
• An interested manufacturer knows they should not say “you need to work on product development on your side, and then come back to me”, as the startup company will get introduced to other manufacturers and the project is not likely to come back to them, so they say “yes, we can support you”;
• From there, things do not progress as planned…

Why? Let’s cover avoiding product development limbo in this article.

Who is this article written for?

If you are working on an innovative product that comes with mechanical features and/or electronic parts, read on. The challenges in bringing such products to market are surprisingly common and similar.

Different capabilities!

You need to understand that product development calls for different capabilities from mass production.

Let’s look at the commercial construction industry. When an office tower is built, the architect decides what to build, the engineers decide how to build it, and the general contractor builds it. These are clearly distinct jobs. I am sure some large companies do several of those activities, but they are probably good at one activity and ‘just passable’ in the others.

Same thing with OEM / contract manufacturers. Their best engineers work on industrialization and manufacturing management, not on new product development. Process engineering is quite different from mechanical, electronic, or firmware engineering.

The biggest factories have a lot of engineering resources and can do a lot of product development work, sure, and you should consider working with Pegatron if they accept your project… but they tend to work with the largest companies.

Since you need to work with a company that gives you a certain level of priority, you need to forget about manufacturers that employ, say, over 1,000 people. And, at that level, they are usually not great at product development.

How to work with a “manufacturing partner”?

The situation really is quite different depending on where you’re at in the new product introduction process. Let’s break it down into three different cases:

Case 1: Giving a manufacturer a lot of engineering design work early in the project

If you haven’t made much progress yet, you may be in the critical phase that we call ‘feasibility study’. (You can see the NPI phases as we describe them here on Sofeast).

Let’s say you have documented the requirements for your product, how it’s going to be used, in what conditions, what would be seen as a catastrophic failure, how it is going to be packed and shipped or deliverable to be sold, installed to customers, and so on and so forth. That’s a start. Let’s also say you have put together some off-the-shelf parts, maybe you have programmed an Arduino module and confirmed the product idea seems to stand on its two feet. Great.

Is it the right time to hand the project over to a manufacturer? Usually not.

We wrote about it last year in Why You Need Mature Product Designs BEFORE You Find A Chinese Manufacturer. If you still have “a sketch on a napkin”, it’s way too early and it leads to this type of issue:

Some Chinese manufacturers may be overly optimistic and say yes even if there are doubts. They skip feasibility and run into problems in their rush to get to production ASAP. They may keep trying to tackle the inevitable problems that will keep coming up for a while, but eventually, their patience wears thin and they drop the project as they tire of wasting their engineering, sourcing, etc, resources. The problem for you, the customer, in this situation, is that you’ve lost maybe 3-6 months and money on the project and you’re still stuck at square one. That’s a disaster.

Case 2: Turning to the manufacturer for prototyping & refinements at an advanced stage of development

If you have already clarified the key components, materials, and technologies, you have a functional prototype that needs refinement, and you have a clear idea of the next steps, this may be the right time to hand the work over.

The question is, are you (at most) two months away from the final validation of the product design? That’s the end of the EVT phase of the NPI process. It involves refinement (for instance, miniaturizing a part of the product, as well as ensuring good “design for manufacturing”).

If yes, then great. You may be able to find manufacturers who will assign competent engineers to finish the job.

Now, are you really sure? Most companies at this stage are over-optimistic.

And this is also true of the suppliers, by the way. They are not likely to tell you ‘oh, your concept still requires a lot of work’. That can be interpreted as a sign of incompetence. If a competitor says “yes, we can do it in two months”, they will probably get the job.

So, are you sure that there is only very minor problem-solving involved from this point forward? Was the design reviewed by several highly experienced engineers? Was there an in-depth risk analysis?

We have seen projects where seemingly a design house had done “98% of the work”, but the remaining 2% ended up in an extra 6 months of development work because just one key issue was overlooked. Imagine a product that uses Bluetooth to communicate with a mobile phone but that is encased in metal, for example… finding a good way to set the antenna up might be very tricky.

The difficulty of estimating the amount of remaining work

Designing a new-to-the-world product involves doing the work, finding problems, then looking for ways around the problems, and on and on. It involves putting together prototypes to test and learn what works and doesn’t work.

How many iterations or prototypes are still needed? That’s nearly impossible to predict accurately for a complex new product.

The problem is, Chinese manufacturers are not known for their patience. If they see no clear progress, their whole attitude changes.

They relocate the best engineers to the next shiny object. They reduce the hours spent on your product. And, after a while, either you take back control over the product development and do all the engineering work on your side, or they would simply drop the project.

They might not say it in such a straightforward way, and the salesperson in charge of your account might be feeling quite uneasy about that. But, in most cases, that’s really what happens.

They all have painful memories of never-ending product development projects, and they want to avoid throwing good money after bad.

This is made much worse when the manufacturer is required to provide a quotation in advance for a lot of the nonrecurring engineering work. If they quote too low and they are too optimistic, they will go ahead until they see that they are losing money, and that’s when things will start to get difficult.

That’s the time they will wonder, okay, is this the last prototype iteration? What happens if the customer rejects it and keeps asking for refinement? Do we drop it?

Then, how to proceed in this scenario?

The best is usually to go with an hourly rate. That’s the only fair approach when the scope of work is not very detailed. Another approach is to work on quotations for the next iteration of design, or a pre-defined block of work.

Anything with a time horizon longer than 1 design & prototyping iteration does not really make sense and may put both parties in a bad situation.

Ideally, the customer works with the development team in an agile manner with very frequent status meetings, a lot of information sharing within small teams, and quick prototyping iterations that include quick feedback from users.

That also requires flexibility in the scope of the features and characteristics of the product. Remember, if you really have to deliver the product by a certain date, you cannot also be very uptight about the budget or about the features of the product that will be launched. The non-essential functionalities may have to wait for version 2.0…

Case 3: Handing the project over to a manufacturer once development is completely over

Now, this might be a bit late.

We have seen products that were supposedly “fully approved and ready to go into production” that were actually impossible to manufacture. An enclosure simply can’t be injection-molded, a surface finishing is nearly impossible to achieve as per the industrial designer’s renderings, and so forth.

That’s why I usually advise companies to get feedback from the manufacturer before they build what they hope is their last prototype. The development team needs to integrate the “design for manufacturing & assembly” feedback from the people who will be tasked to develop & validate the manufacturing processes.

What is perhaps the approach least likely to succeed?

I have yet to see a project that combined the following two approaches and went into production without plunging into product development limbo and suffering from massive delays:

• Requiring the manufacturer to handle most of the development, including acting as the tech lead
• Requiring an offer (with commitments about money and time to spend) upfront for all the development work

I understand the temptation to go down that route. Some companies want predictability. They have a product to develop and need to set up a budget. However, the temptation to go out to the market and request quotations will usually provide a false sense of security for the reasons I explained above.

You realize you have made a mistake and are wondering which way to go.

That’s pretty common, and it’s better to come to this realization late than never. Once you have sent payments for tooling, switching to another factory is much more difficult.

If you are in product development limbo, I’d urge you to read Switch Away from a Manufacturer at the First Signs of Trouble.