API – Always Pushed Indefinitely?
Digitalization of the electronic components supply chain is the way to go. That is a very common agreement among different players in the industry, such as OEMs, EMS providers, distributors, and OCMs. But does the agreement end there?
Application Programming Interfaces or APIs–connections to exchange data and manage transactions between the above-mentioned players–have existed for a while in many industries. However, one of the most innovative sectors of the global economy, the electronic components business, handles too many of its activities the old-fashioned way. This leaves productivity gains as well as better and faster information exchange on the table.
Reasons for a broad reluctance may be manifold, but so are positive examples of where APIs work. We tried to figure out a bit more about the status of behaviors in the industry and the potential for APIs by doing a series of polls on LinkedIn. The results are by no means representative but a good indication of the current state of affairs within the industry.
Purchasing behavior
The first poll question read: which of the following is the best way to purchase electronic components? Results were only kind of surprising: one-third of voters favored API (32%) as the best route to do purchasing, followed by the traditional pathway through sales reps (27%). EDI, once the classic “digitalization” must-do, was favored by less than a quarter (23%). Websites, specifically online orders, are the weakest choice with about 18%.
As we didn't distinguish between volume and online/catalog distribution, I will take a bold guess and assume that the 18% either represents the share where online makes sense, small volume, fast delivery, or it does not give a customer what they would want across the board yet. No surprise.
However, buying through sales reps is still seen as being a very favorable way to purchase components: long-standing relationships, the human factor, trust, and the ingenuity of salespeople to find solutions for their clients. The real surprise is the fact that one-third mentioned APIs as the best way to purchase. It sounds like a clear understanding of the potential of APIs in key transactions like – orders. The open question may be this: what is the best way to buy today or the best way in the future?
Status of API adoption
One remark upfront: there were very few votes, and any interpretation would reek of over-assumption, therefore just a few words. The extremes (little adoption, full strategy already in place) took the fewest votes. The middle ground–trials with third parties and strategy ready to deploy–seems to be where people or others see the industry at large.
Knowing the market, I would assume that the things that are going on at the moment are far from being perfect and mostly limited to the simpler APIs--Data Gathering, Price & Availability, mainly with standard prices--as these are the only ones easily transferable through an interface.
The constraints, especially when it comes to quoting custom prices, for example to a contract manufacturer who might have a dozen different contract prices at a distributor for one and the same part, prevent an easy solution. Further, customers may also have limited IT resources to transfer such a work-heavy process into an orderly digital process.
To summarize this part of the poll: More enlightenment on the benefits of API needs to happen, clearer communication on API roadmaps (at distributors), and, last but not least also, further investigation into customers’ pain points and where APIs can help them.
Types of APIs with the most impact:
The question simply read “Which type of API do you think is most useful to customers’ transaction efficiency?”. The choices were between data & part search, offer (price & availability), order (placing an order from your own ERP through the API), and market intelligence & analytics.
As indicated in the other polls, please do not expect statistical relevance, just a trend. Offer API made the top spot with over 50% of votes, followed by Part Search (23%), Market Intelligence (15%), and Order (8%). This looks like a fair representation of what is in place today, either directly to manufacturers, distributors, or platforms that are used in the marketplace--let’s call them Quote and BOM switchboards. It corresponds also with the second poll where trials with 3rd parties and ready-to-use API strategies were favored (caveat!).
However, compared to the importance that is assigned to APIs in general, this seems to be a clear shortfall of expectations. Please consider in this context that the so-called Offer API rarely produces a binding quote, as price and availability are based on general system data not yet necessarily on custom or project prices based on manufacturer quotes. If you got that, you are close to an Order API and would do it directly from ERP to ERP rather via a switchboard. In other words, the two favored APIs – Offer and Part Search – indicate early trial rather than a full strategy figured out. The full monty comes with Order API.
A few more assumptions from trend data focus on the stage of maturity, not of the customers, but of the API world in the electronic components business. Order API is something very few companies have in place today – like Sourceability and some distributors with only selected customers. If I had to guess, I would pick one leading analog manufacturer and two global catalog distributors to be in the lead in this category--with some relentlessness, for sure.
Market Intelligence APIs, as much as they would be needed, currently lack the data intelligence analytics power to be relevant with one exception: Sourceability’s Datalynq market intelligence tool with its predictive analytics function. However, adoption so far is more as an online Software-as-a-Service-Method than as API.
Market intelligence, specifically data analytics in general, shares a common challenge: quality and completeness of data and lack of good algorithms to turn information into real analytics or, more preferably, good predictions. Customers today very often struggle with the cost of simple data such as parametric and material data, unable to even consider the larger price for meaningful predictions based on real transaction data. Once you include future challenges around dangerous substances, data providers charge premiums and still have many holes in their databases.
What about Order API then? It’s working, but adoption is very low and conviction about the positive impact on their own transaction efficiency--order management--is not very high among customers either. Not yet at least.
It is a mix of dependency concerns, how many suppliers do I have to have to need an API, and fear of change, mostly in purchasing departments about job roles, that keeps adoption at bay. These concerns combined with a lack of IT resources to get it done, such as ERP to ERP exchange of data and transactions. This is not a complex requirement but needs strategy and an exact understanding of data mapping. From my own experience, there is always another IT project in the way of getting something new done. Sound familiar?
APIs will come, so too will a more transparent and efficient–hopefully global–sourcing process. Not because they will be the ultimate solution to all problems and mismatches in the supply chain, but because they are the first step in identifying and minimizing all the existing pain points in the supply chain. APIs will finally do what needs to be done in a low-margin environment: automate manual processes and drive needed efficiency.
