In this blog Contributing Analyst Barry Graham, author of a new TM Forum report called Full stack transformations: Do they work and is there an alternative
, looks at how the procurement model is changing – for the better.When I first started working for a major telecoms equipment vendor, I remember being surprised at the scale of customer interactions. We would routinely send a team of five people halfway round the world to meet with an existing customer, and that was nothing compared to the lengths we would go to secure a new contract.
At the time, of course, it made sense. We were spending the cost of a medium-sized family car on one meeting but winning contracts that were worth the cost of a commercial airliner in return. We rounded things to the nearest $1,000, so if we wanted to put a cost of $100 in a spreadsheet, we wrote it as $0.1K. That was the scale we understood, and the scale our organisation was optimized for.
In researching the full stack report, it became clear that the advent of modular architectures is driving a big change in the relationship between communications service providers (CSPs) and their vendors. One thing quickly becomes clear: the age-old request-for-proposal (RFP) process, with a procurement team focussed on getting the required functionality for the best price, is no longer a viable model for the kind of technical partnership needed to effect a modular transformation. This is an area my colleague Mark Newman researched extensively in 2019 for the report
Time to kill the RFP? Reinventing IT procurement for the 2020s.
Time to downsize
The other thing that struck me is that the size of packages CSPs are buying is taking a significant step down. In a modular IT systems such as those described by the
TM Forum Open Digital Architecture, a piece of software that was once a single, full stack solution is now broken down into perhaps four or five decoupled components.
The CSPs I spoke to told me they are now moving to procuring software at this component level, mixing the best from different vendors. At the moment they are still working with a small number of strategic partners, so perhaps the total spend for those relationships has not changed very much yet, but we can clearly see the direction of travel.
If we take componentization to the extreme, to a true microservices based architecture, we are not talking about a handful of components. We are talking about literally hundreds of microservices. What if we start mixing vendors at that level of granularity? What if we move away from capital purchase to a software-as-a-service business model?
On the price lists for Amazon Web Services, many of the prices for their services are stated to four decimal places, that is, to the nearest one hundredth of a cent – or $0.0000001K in the notation of a 1990s telecoms vendor.
Unanswered questions
Now, nobody is telling me they are even thinking of procurement at the level of a microservice at the moment. The complexity seems to outweigh the benefits, and it is not clear what the optimum-sized unit will be in the 2020s.
Whilst I suspect it will be bigger than a microservice, I am very certain it will be smaller than anything we are used to working with today, and that is going to represent a seismic change for the commercial departments of the traditional vendors. As ever, a seismic change for the incumbents represents a huge opportunity for new entrants who may well have businesses far better optimized to work at the scale required, and certainly there is appetite from the technical teams I spoke to for exploring new innovative vendors.
Change is the only constant.
