Predictive analytics software for manufacturing: build or buy?
You know better than anyone how unique your company or facility is. It can feel tempting to create a software system catered to your specific needs and use cases — something that uses your own internal terminology and has the functionality that you know you’ll need over the long term. Maybe you’ve gone as far as exploring some analytics solutions on the market, maybe you’ve even sat in on some product demos and you have decided that no external platform is quite right for your business.
This is the same thinking that some years ago prompted some companies to resist using external software for things like Human Resources, Accounting, and Sales… Many organizations spent hundreds and even thousands of hours to create their own versions of standard tools for internal use. I’m sure you’ve read about companies that tried to build their own CRM only to frustrate employees and eventually purchase an industry-standard tool a year later.
Software categories have matured and many competing programs and platforms are now offered. Today, it would seem foolish for most companies to develop their own Human Resources Information Software when they could so easily use something like Ceridian or Paycor. And, no one would try to build a CRM when they could easily use something like Salesforce.
If those companies had trusted the development of this software in the market and waited a little longer, they could have saved enormous internal effort spent planning, hiring specific talent to build, building, and maintaining these internal tools that almost always end up replaced.
There are some situations that do warrant building an internal software system, but we would argue that these situations are rare. The global business software and services market was valued at $474.61 billion in 2022 and is expected to grow by 11.9% from 2023 to 2030. As a result, there is software available for almost any business use case you can imagine, and even some you can’t imagine.
Twenty years ago, no one was thinking that there would someday be a tool to help you predict when a quality issue would occur in production. The idea that you could take action to prevent a pending problem proactively would have seemed like a far-off fantasy.
Today, it’s a reality. With products like LinePulse on the market, why would an automotive manufacturer consider starting from scratch to create something so complex?
Why you shouldn’t build your own software for manufacturing
Here are some of the top reasons why we think precision manufacturers should not attempt to build a software solution:
1. Software development is outside of your business’s core competency
Businesses succeed by focusing on what they do best. Companies like Apple, Amazon, Tesla, and McDonalds have all succeeded by building their businesses around the core attributes (not necessarily the products) that make their companies uniquely competitive. They know that splitting their focus can lead to disjointed management and operational confusion. Juggling multiple priorities leads to inefficiencies and distractions, moving your people away from what they should be executing on.
A great example of businesses that have taken on software development in addition to the core part of their business are banks. Most traditional banks developed their own IT systems and software while that industry was beginning to undergo its own digital transformation. However, one could argue it hasn’t been overly successful. Banking IT systems, apps and software tend to fall behind the rest of the tech sector in functionality, ease of use, and design. Often, layers of systems were built on top of each other over time that have gotten convoluted and complex, making it challenging for new employees to learn how to manage them. Old code bases that lay the foundation for these systems are outdated, and small issues can be hard to fix without bringing the whole leaning IT tower down.
2. Internal software is developed by smaller teams with limited capacity
The key to creating a successful piece of software is all in the team. You may have a robust IT and development department already, but creating a new product is an entirely new beast compared to ongoing procurement, operations, and maintenance work that most IT teams in manufacturing would already be doing. Is this team thinking about long-term business needs? Do they have the skills to design, develop, test, deploy, and maintain a new product?
If you don’t have these people in place, you will need to hire a few. In that case, does your business have what it takes to attract top talent in the tech industry? Successful candidates are used to high salaries, flexible work arrangements, and a specific style of work and management. What extra work needs to be done to attract and hire the right team to execute this software build? How much will it cost just to get that team together?
Additionally, once your all-star product development team is assembled, it is likely that this team will be rather small compared to the larger ones at many software development companies working on analytics platforms for manufacturing. A small team can work in a fast and agile manner, but what often falls to the wayside is proper documentation and procedure. Within small, busy teams, excessive documentation is not necessary to do the ongoing work. What happens if two of these project team members leave in the middle of the development process? How easily can they be replaced, when they are taking months or years of domain knowledge with them?
Even after the software is successfully deployed, is there a plan in place for what to do when turnover happens in the team and the original developers are not present to help when problems arise? What about long-term updates and adding functionality to the software?
3. Developing software internally is expensive
Let’s look at the costs that may be involved to develop your own platform.
Taking an idea from concept to inception can be quite a daunting task. Leaving out all ancillary spend in finding, training and managing your software development team, a simplified example can be illustrated below.
Company X building their own analytics platform:
- Consider a team of 5 software developers (UI, UX, Data aggregation, Code, ML algorithm)
- Average salary: $70-$150k annual = roughly $110k per developer
- Total employment cost: 5 developers x $110k = $550k/year
- Software development timeline: 1.5 years (Assuming a perfectly timed, perfectly executed rollout)
- Total labour cost to build platform: $550k x 1.5 = $825k spend
Buying an off-the-shelf software solution:
- Total cost: $50k annual subscription base
$825k / $50k = 16.5 years to reach breakeven
Of course this math is simplified, but it does demonstrate the wide gap between the cost of developing analytics software internally, and purchasing a software solution. The lower cost required to purchase software means that you can see your ROI much faster.
4. It can distract from growth in your core business
The software development process is very different from that of manufacturing physical products. There are different project management styles used, different ways of reporting on progress and efficacy of the product, and different expectations by employees that work in tech. Upper management can get caught up in the progress of these side projects, and by splitting their time and financial resources up, a little less will be accomplished in the core business.
Speed to market is getting more and more important in the automotive industry, and this means that the level of focus necessary on key products and product lines is at an all-time high. There is little time to waste on other endeavors.
5. Complex software solutions take a long time to develop
The time that it will take your business to get a software development project approved, put the right team in place, and begin the planning and design process could be a year or more. Not to mention the time it will take to design, develop, test, and implement the software once it is finished. Oh, and did we mention the need for ongoing support and maintenance?
During the few years that this entire process takes, other software companies out there are refining and adding more features to their tools. Perhaps by the time you finish developing your custom solution, you realize that there is one already on the market that has the same functionality you were looking for in the first place.
There is also the opportunity cost of the years spent without using any analytics software at all to consider. What money could you have saved, or efficiencies you could have found using another tool during that time?
Benefits of using an external vendor for your predictive analytics software solution:
1. Larger, dedicated teams create better software
This is the big one. When a company is entirely built around designing software, they have worked to hire experts in the industry and highly skilled employees to create their products. The management team and employees are tuned in to the trends in software development (for example, the different ways to build machine learning algorithms for a predictive analytics solution), and they are keenly aware of issues like security, maintenance, long-term stability, testing, and technical support. They know what it takes to bring a successful piece of software to life, and how to do it efficiently.
These teams have often created software before. They know what a product roadmap should look like, and how to overcome obstacles in the process. They are also planning extensively to make the tool as easy to onboard and as easy to use as possible.
They are thinking about the long-term use of the tool in manufacturing, and how and when new features should be added. They likely have teams available to train new users to use it, and technical support working to resolve issues. Their reputation as a company depends on successful deployments and repeat business.
2. External analytics software stays updated and cutting-edge
Tech is highly competitive. Once a unique product emerges, it isn’t long before competitors pop up. Software development is a fast-paced industry that uses speed as a major advantage to compete in the marketplace. Software developers are not limited by physical constraints such as building infrastructure or even having a fixed office location to create products. This makes it possible for companies to be as agile and quick as their developers can code.
Like any competitive industry, tech companies keep a close eye on their competition. To stay relevant, they must evolve to do the same thing their competitors can do – and more. Using an external piece of software from a successful and reliable source ensures that it will be a cutting edge tool, because that is the only way a successful software company can stay in business.
Another thing that drives software companies to improve their tools is user feedback. When they have requests from their users to add or change certain features, it is in the company’s best interest to integrate this feedback. Other manufacturers might have insight into how to improve these tools that you can benefit from. If you create your own tool internally, you are limited by only the features and designs that your team can envision.
3. Flexibility to change platforms more easily
If you are already using an external software solution, it is easier to swap to a competing platform than it would be to swap from an internal software solution to something else. The time and cost put towards developing software is only worth it if the software is used over a long period of time. After putting the work in to developing something internally, there is strong pressure to see it through, even if the end product is less functional or limited compared to what is available on the market. Conversely, if you buy your software solution, you have an opportunity to test out a fully-designed platform with little upfront cost.
Additionally, once an external tool is used, it has already been proven that your facility has the necessary infrastructure and outputs to use an external piece of software. Perhaps some initial setup steps need to be put into place to upgrade your systems to make using an external tool possible, such as safely connecting to the cloud. Once these steps have been taken, it is much simpler to switch to a competing product.
4. The ability to use OpEx instead of CapEx for external software
For accounting purposes, developing an internal software system is likely to be a long-term investment put towards the Capital Expenditures (CapEx) budgets. CapEx is reserved for long-term investments in fixed assets that eventually depreciate over time, such as buildings, trucks, and machinery. Getting an expense approved as a Capital Expenditure can be a lengthy process, as the business needs to be sure that the asset will offer long-term value.
When you purchase external software, especially a SaaS solution with an ongoing subscription model, this comes out of your Operational Expenditures (OpEx) budget. OpEx are business expenses like employee wages, rent, R&D, legal expenses, etc. Operational expenses are considered short-term, and are used up within the year they are purchased in. For example, one year of rent or one year of employee wages.
By purchasing software under OpEx, the budget approval process is usually faster and easier. When the software becomes an operational expense, it offers more flexibility to the business, and is easier to manage from an accounting perspective.
After all that, I still need a custom analytics solution
If this is where you are at, there are still some options for you that don’t require building your very own software solution. Software companies are eager to build helpful products for their large clients, and often work in partnership with manufacturers to tweak or customize off-the-shelf offerings for specific use cases, or even build a custom solution. If you need a predictive quality analytics software solution tailored to you, get in touch.
Share on social: