Standards & Intermediation
Part 2 on how logistics platforms compete
This is the second in a series of articles on platforms as a business model for the logistics sector. It connects with a three hour podcast in which I discussed the same topic with my friend Märten Veskimäe. Check out the podcast at Logistics Tribe on Apple or Spotify.
In 2020 I was rebuilding my company’s website (www.tnx-logistics.com). Our product was fairly contrarian: we had a perspective on what the future of spot procurement was about and our product invited you to become that kind of company through non-trivial transformation. The more I reflected on this aspect of our product (and the team who built it) the more I knew we needed aesthetics on our website to match. I wanted a website visitor to have an immediate, gut-level reaction that this website was built by people who care a lot about visual experience, who think abstractly, and also aren’t afraid to bet on our expertise. We engaged Rafael Verona, a visual artist who was pushing the bounds of what can be done with micro-animations. Here is an example:
The fact that his final art was nearly impossible to make work on the wide variety of devices and bandwidths of our visitors was secondary to the pure aesthetic signal. I loved it.
But we also had a LinkedIn page: cookie-cutter visuals and experience. Given how much our team invested into our website, why would we tolerate being so constrained? Here we get a lesson in platform value through standard setting. While we could all use Google search to find the webpage of people and companies, this loads much more effort onto both sides. It is harder to make a website than to fill out a LinkedIn profile. And it is much more mental energy to read through a myriad of designs. So, we all accept the constraints of LinkedIn to cut down on our work to present ourselves and also our work to find and understand everyone else’s profiles. Setting the standards by which members will interact is a primary focus of a platform.
I'll zero in for the rest of the article on platform-to-platform competition, but as a recap I think platforms compete with three groups:
Platforms compete with non-platform business models. For example, taxis compete with ride hailing platforms. Instead of facilitating interactions between counterparties they become the counterparty directly. These competitions tend to come down to risk of failed interaction. A direct counterparty can absorb risks in time (i.e. liquidity) and execution quality (i.e. liability) in a way that platforms will not. But direct counterparties expect higher fees for these additional services. In this sense a platform looks like a de-bundling of the value-adding services of a direct counterparty. A platform like Teleroute will enable matches if the counterparties can and want to do so. But they need to be willing to deal at the same moment and they take full risk against each other. Teleroute is cheap compared to a forwarder, who would also match a shipper to a small asset owning carrier but also take time risk in committing to rates with one before the other, and who is legally liable for execution or payment.
Platforms compete with their own members. For example, if I use Jaegger to conduct my logistics procurement events there is a constant potential that I decide I’d rather work directly with my suppliers. The platform is an intermediary, after all, and it can be dis-intermediated if low unique value & switching costs suggest doing so.
Platforms compete with other platforms. Yes, but how? In the first article of this series we looked at the over-focus on scale as the only dimension. Platform owners may also say it’s about features & functions: I’m skeptical. Even great feature advantages are short-term advantages. Competitors will mimic or counter-position. I’d contend that standard-setting is one of two ways in which a smaller scale platform can unseat a larger scale one when all other factors are only negligibly differentiated. The other way is to have a higher proprietary role on the match or interaction.
Bonus Competitor: Platforms may someday compete with tokenization. There is at least a whole article’s worth of inspection on how tokenization and web 3.0 competes with platforms. But for now we can say that this is similar in nature to how platforms compete with other business models. If a forwarder gets price discovery through acting as the counterparty, and platforms gets it via being an intermediary, web 3.0 does so via community-enforced protocols. For now there are simply no meaningful web 3.0 competitors to the major logistics platforms, so not much need for competitive analysis. Maybe someday.
Competiton on Standards
For members of a platform the enforcement of standards reduce friction to make the desired interaction. It simplifies cognitive load on user, both the upfront work of how to communicate or workflow through the interaction or the transactional effort to communicate each time an interaction is underway. Standards also cut off types of risk one would otherwise have to insure or avoid through extra effort. Standards are a bit like the rules of a highway, with its speed limit and one-way driving, and rules to stay in one lane. These make it easier for us to drive (easier to learn, easier to do it around new places & people), and they reduce the risk of accidents. Taken together rules of a highway enable more traffic through it. Everyone wins.
But the downside to platforms is that each rule boxes in possible behavior in some dimension: on the highway one cannot drive very slow, straddling lanes, into oncoming traffic. If we look at successful platform competition where an underdog overtook a leader, a major tactic is to have restated the standards in a way that enables paradigm-changing (and eventually better) member behaviors. As an example: WhatsApp structured their member creation so that your phone number is your identity, but not the method of messaging. No need to have an email (like Facebook) and no need to use the mobile network for messaging (like SMS). It turns out a lot of people wanted to socialize without having an email, essentially.
Why wouldn’t a novel standard from a rival platform simply be copied? Maybe they are copied, but in many cases I presume that standards are deeply embodied in platforms at both a code and member expectations level. For certain standards, they may be so fundamental that they negate scale advantages. For example, I am personally skeptical about the “huge datasets” of loadboards like DAT, Timocom, etc because I know they have low data quality. If the noise is ever going to be separated from the signal, one must do it early. Downstream cleansing is a desperate thing. Platforms like DAT or Timocom that do not enforce high degrees of data standards effectively trash their data in real time. If they were to increase their standards on data quality, they would never be able to retrospectively do so for the past transactions. And for this reason, I think a novel standard (that leads to higher data quality but lower scale) on a competing platform would negate the scale advantage (more data of less quality) in a way that even a fast-follow doesn’t resolve.
My sense is that standards are a super interesting aspect of platforms in that they probably overtake each other in the way of paradigm shifts in science or art. They would appear to be almost incommensurable, and that’s what makes the ability of the incumbent to follow a newly emerging standard so challenging.
Competition on Depth of Intermediation
Platforms may not act as counterparties, but they can take a shallower or deeper role in the intermediation of members. I think this is the third meaningful way platforms compete with other platforms, after scale (discussed in the first article of this series) and alternative standard setting. In my mind I categorize intermediation into four depths depending on how the sides of platform interact. Here they are, starting with the most shallow and going to the deepest intermediation:
Navigation: this implies the sides know who or what they want to interact with and the platform just handles administration details.
Search: this implies that a member doesn’t know who or what they’ll interact with, only what they hope to find. The platform must structurally comprehend its own content, but the member knows their own satisfaction criteria.
Recommendations: this implies the platform knows the desires of the members as well or better than them, plus the platform catalogue, and attempts to initiate an interaction. In this situation the member wouldn’t necessarily know how to formulate their own search, or be clear about their own satisfaction criteria.
Assignment: this means the platform can recommend but also goes further executes the interaction. Think how a rideshare driver is assigned: platform does it, not me.
My expectation is that at each level it becomes harder to compete with the platform because more and more of the work for a successful interaction is the responsibility of the platform (and therefore their intermediation has high stickiness).
Summarizing Platform Competition
Across two articles in this series I’ve described how I see competition for logistics platforms. Platforms serve some set of members as they seek interactions with each other. The platform is enforcing structure and going to some depth of intermediating the specific interaction. As an intermediary, the platform competes with dis-intermediation (i.e. the members of the platform can decide to just work directly together). Platforms also face alternative models of intermediation, namely a traditional counterparty, and perhaps someday in web 3.0 some form of tokenization, smart contracts, distributed ledger, or other self-enforcing protocol used by community agreement. And, of course, platforms compete with other platforms.
The platform-to-platform competition seems to turn on three differentiable points. The first is the most obvious: scale. Scale can indeed create competitive differentiation if the platform can prevent multi-homing, indicates member diversity, and is dealing with highly heterogenous matching criteria. Scale always offers diminishing returns, but a platform that creates additive network effects (rather than swapping effects) reaches its plateau of scale-value at a later point.
While scale is the most discussed platform-to-platform competitive dimension, the other two still deserve attention. Platforms define the standards for their community of members, and in doing so they create a method by which one platform can out compete another regardless of scale. Standards as a means to compete is one of the more exciting aspects of building and running platform businesses because it allows for counterpositioning and therefore an upstart winning against an incumbent.
Finally, platforms compete by offering differing levels of intermediation, ranging from simply enabling the navigation and selection of members to work with through recommendation and even binding automatic assignment. Each level raises the stakes on what would constitute success for the platform, but also makes the platform that meets those expectations much more sticky.
Background Research
If this discussion does something for you, consider also checking out these sources that I found useful.
The Platform Delusion: Who Wins and Who Loses in the Age of Tech Titans
7 Powers
Acquired podcast: Platforms and Power (with Hamilton Helmer and Chenyi Shi)
The cold start problem
Platform Scale
Matchmakers: The New Economics of Multisided Platforms
https://hbr.org/2019/01/why-some-platforms-thrive-and-others-dont
https://www.hbs.edu/ris/Publication Files/18-032_d71914fe-d56c-42ad-ae20-deb5b979fab9.pdf
https://op.europa.eu/en/publication-detail/-/publication/57c7d6ed-ad52-11eb-9767-01aa75ed71a1