A surprising number of prominent French companies attempt to develop custom visualization technologies when mature commercial offerings already exist. This article shares observations explaining the reasons why.
In the late 2000s, I ran a management consulting startup. Our team specialized in helping companies streamline internal processes and implement enterprise resource planning (ERP) platforms with optimized capabilities for specific business types.
At that time, the adoption of on-premises ERP systems among small and medium-sized businesses had reached its high point, as many companies in Ontario and the northeastern United States transitioned away from archaic legacy systems (often involving paper, spreadsheets or applications created for mainframes).
One industry stood out among the countless businesses adopting ERP technology as the technological backbone for their organizations – the Architecture, Engineering, Construction (AEC) industry.
AEC’s Unique Concerns
Our team had a difficult time in the ERP selection space with members of the AEC industry. Our uniform feedback confirmed that the main culprit was the unique philosophy that guides many AEC firms. Unlike their counterparts in manufacturing or professional services, construction companies were reluctant to accept commercial ERP offerings. Instead, many AEC firms pursued the in-house development of custom ERP systems.
The top three reasons behind this unique approach were summarized as:
- Product fit
Cost was often cited as the key reason for pursuing in-house development. Whether commercial product implementations ranged from $0.75M to $5M, internal development was often viewed as a cost-efficient alternative. In addition to the high list prices of ERP software components, an integrator company typically charged between $150 and $200 per hour for business and technical analysts to install and configure commercial packages. The hourly cost of an experienced in-house developer with managerial and other overheads was less than half.
Self-reliance was another key consideration. Integrators have multiple clients to support, and their knowledge of individual client needs may deteriorate over time. Relationships with external parties may evolve but not always for the best. Internal teams knew how to support the enterprise at a much lower cost and with better results.
Finally, greater product fit can be achieved with systems designed by the company specifically for its own needs. Commercial solutions may not have the flexibility to accommodate all business specifics, even with customizations.
This was the guiding philosophy. In practice, it couldn’t be further from the truth.
Extraordinary costs, amateur quality
When I caught up with my former industry associates, I learned that many adamant ERP sceptics eventually scrapped their grown systems in favour of commercial solutions.
First, the cost – maintaining an internal team turned out to be a much costlier endeavour than estimated. While commercial solutions offered proven products that required minimal (and predictable) configuration to fit the company’s processes, internal systems had to be built from the ground up. Labour requirements for the designing, building and testing of a system were substantially greater than tweaking an existing solution designed specifically for the industry, even if that home-grown system represented a fraction of the commercial system’s functionality.
Self-reliance proved to be illusionary too. Employment for a lifetime became rare in the 2000s. As attrition took its toll, companies found that the only employees capable of maintaining their home-grown systems eventually left, and their replacements were unable to attain the knowledge needed to support them. While ERP integrators had deep talent pools of trained specialists, in-house systems were black boxes that would take years to reverse-engineer and learn.
Finally, product fit expectations were not met either. Shifting and conflicting priorities shaped products that offered poor functional fit. User Experience and User Interface is something that takes years to hone, and most internal teams have limited experience in designing consumer-ready products. The resulting systems were often awkward, with confusing workflows and terrible usability.
While seeking cost-efficient, customized, smooth-running productivity tools, companies ended up with extraordinarily expensive, productivity-killing monstrosities, and the only employee who could keep them on life support had just left the company.
Let’s fast-forward to the late 2010s, when I was with an XR startup. XR – or extended reality – refers collectively to augmented reality (AR), mixed reality (MR) and virtual reality (VR) technologies. This area promised to reshape many industries – from transportation to manufacturing to healthcare. AEC companies took notice and began experimenting with visualization tools.
I must admit that I expected AEC companies to apply an ERP-like adoption approach to the XR tech. I was surprised to discover that the industry’s philosophy took a 180-degree turn. Most AEC companies fixed on identifying suitable commercial tools rather than building their own. This trend was global; the vast majority of companies in North America, Western Europe and Asia-Pacific focused on building portfolios from leading commercial offerings.
With one exception – France.
France has a vibrant startup community that produces exciting innovations. It was not surprising that at the onset of XR in 2017–2019, French companies showed early interest in advanced visualization tech. However, much to our surprise, we have seen the same companies that were scouting trade shows for commercial systems turn to building home-grown applications. Social media and occasional articles displayed rough prototypes, with groups of proud executives shaking their heads in the background. Articles trumpeted groundbreaking technologies that would one day change industries, while being completely oblivious to the existence of commercial tools that had been doing the same job for years.
We interviewed our French associates and employees of the companies that had turned to in-house development and identified three factors behind this approach:
- Product fit
Cost was cited as the justification for investing often shocking amounts into internal development or hiring third parties for contract coding. Two companies that invested eye-watering sums justified their investments with simple math – if for every employee in the company (thousands) the firm buys a commercial license (hundreds of Euros a year), it will cost millions every year. If a year or two of such subscription is invested in bespoke software, it will be a one-time cost with benefits lasting years.
This math is faulty on multiple levels – the development cost is understated while the subscription cost is exaggerated. In a realistic deployment scenario, the cost invested in custom development pays for 10–20 years of the full organization’s subscription needs. It is the same reason why companies pay $25 per employee per month for Office 365 subscriptions instead of spending hundreds of millions of dollars to develop proprietary mail systems, conferencing solutions, spreadsheets and word processing tools.
As for self reliance, engineering prowess is rightfully a point of pride for many French companies. In this context, participation in trendy developments is sometimes mistaken for innovation. Evaluating AI or block-chain to deliver better services is different from designing a custom block-chain-powered AI platform with an unclear purpose or business value. The former is an example of innovation; the latter is not.
In a highly globalized world, companies focus on core competencies while sourcing other components from those who specialize in making them. Car manufacturers don’t design CPUs for their cars, just as aircraft manufacturers don’t formulate rubber for aircraft tires. This largely applies to AEC companies too – they don’t design computers for their engineers, just as they don’t manufacture excavators, cranes or cement mixers. However, in the XR area, a surprising number of large French companies invested in internal XR prototypes when mature commercial offerings already existed.
Self-reliance in a niche area with no relation to the core business is illusory. The scale of internal development is unlikely to create a world-leading XR development powerhouse, and institutional knowledge is likely to be lost over time through common attrition.
Finally, the product fit. Just as with the ERP situation described earlier, custom development doesn’t necessarily translate into better product fit. The data residency requirement is often cited as a critical reason for pursuing local development. However, as many commercial solutions can be hosted in different regions to comply with data residency laws, this argument to justify investments of hundreds of thousands, if not millions of Euros, falls flat. Interviews with end users from companies that developed internal solutions reveal disappointment with the chosen direction. The general feeling is that internal politics forces the adoption of products with substantial capability gaps, while much more capable products remain out of reach.
Innovation vs. reinvention
The willingness of multiple industries to adopt advanced visualization tech signals the understanding of corporate leaders of trends and technologies likely to define the future. We see it all over the world – corporations began deploying pilot projects with Microsoft HoloLens in 2017–2018, expanding them to cover phone-based AR and in-office VR pilots in 2018–2019. Many projects went down dead-end paths. However, many led to unearthing use cases and drafting roadmaps for company-wide deployments.
The unique approach taken by many French companies syphons funds away from R&D and innovation. While their European, American and Australian peers build expertise in using and scaling XR deployments, these French organizations focus on reinventing products that already exist.
There are instances where internal development is critical. Geospatial industry leaders – including Bentley, Cesium and Esri – developed data-driven technology forecasts. They invest in creating components to support their core products in the context of future customer needs. However, companies that pursue development due to inadequate market research or the “cool factor” of creating a custom high-tech solution are unlikely to succeed. Instead, they seem to follow the path of those who tried to create their own ERP systems, which led to costly internal development followed by years of struggling with productivity-killing systems and the eventual adoption of commercial solutions.
Opinion by Alec Pestov.