On CSMS valuation hype (and why you might roll your own)

This article is broadly concerned with large corporates having CPO ambitions; a recent post (copied below) on Charging Station Management System (CSMS) vendor valuations was coincident with a few parties asking whether it's likely we'll see more indigenous efforts to build and/or own a CSMS platform. On the latter at least, I hope so. More below.

Origins (maybe)

See this picture here?

The genesis of this kicked off ~15 years ago by a coalition of Dutch electricity distributors and went on to define the industry. The blue dots are major stakeholders with key functional roles. Anyone locally familiar with the DNSP Prime Directives on keeping networks intact can almost imagine how this went:

- "Obviously we want EV charging to integrate into the grid in a controlled way, yes?"

- "Ja, akkoord" (yes, agreed).

- "OK then, so we have the DSO over here, see? Good. So then... retailing?"

- "Nee, dat is niet onze rol" (no, that's not our role). "Zet daarvoor een blauwe stip op het diagram die aangeeft wie dat doet" (put a blue dot on the diagram representing whoever does that).

- "OK then... owning all this new infrastructure?"

- "Ah... Nog een blauwe stip..." (ah... another blue dot...)

This would have gone on a while.

(Whatever the actual process) it ended up creating a blueprint the industry still uses today to identify roles, enabling standards/frameworks/protocols, per market/geography substitutions and (critically, importantly) boundaries for interoperability. From a functional perspective it is terrifically well-considered and continues to guide the development of frameworks and standards making better charging possible. This approach was developed at a time when the ways to communicate between an EVSE and EV were elementary but standardised, when the earliest CPOs already existed with no means or intent of interoperability. It immediately gave us the Open Charge Point Protocol (OCPP) which has contributed greatly to consistent functional and investment efficiency expectations around vehicle charging infrastructure and related experiences.

And from a go-to-market perspective can be a relative nightmare - let's just say it and exhale - there's no industry in the world that, by design, requires so many disparate stakeholders to get their sh*t together in a new technology sect as interoperable vehicle charging. None. And that's just to get something conformance-functional out the door - if you want it to actually work and be winningly-competitive in market, those players need to work together, all rowing in the same direction, and have similarly low-margin views about their contributions.

To be fair it is a very necessary nightmare and the industry does very well to promote efforts to have all work together:

• Necessary (in part) because interoperability is genuinely hard but genuinely necessary; there is one Tesla and replicating as much across vehicle OEMs would simply create a potpourri of walled gardens, which is no good for anyone and which limits investment efficiency in the space considerably (it's easy for armchair commentators to coo at Tesla's connector and suggest it as the cure to all ills; what lies behind is what makes Tesla's charging experience great however zero industry interoperability diminishes the challenge and makes some aspects considerably easier to optimise), and
• Both CharIN and the Open Charge Alliance regularly host events where vendors can test this much in an industry-collaborative way. I've been and they're excellent events (if a long way from Australia).

Real concerns are more on the commercials; let's dig in.

CPO love

The basic premise of a CPO is pretty straightforwards:

• We will own stuff (in this case, charging infrastructure),
• We will rent that stuff to those that need it in order to obtain recurring revenues, and
• Where possible we will try to mop up/integrate with opportunities that appear broadly similar, manageable and have recurring revenues.

That last point defines why a number of CPOs angle after an eMSP role - energy retailing doesn't seem that hard (ha!) and for utilities, owning infrastructure feels familiar (double ha!). There's additional margin, the unified front can create customer stickiness and any analyst-and-up Big Four consultant on a short PowerPoint deck of 'things your large corporate could own and make money from' will point out that ultimately EV loads could be self-firming - as in there may come a point in future where EVs make it easier to be a viable energy retailer.

These suggestions detailing a glorious corporate march towards fiscal greatness tend to run out of bravery when talking to the tech bits; to be fair, most upper management corporate audiences used to infrastructure and/or services may have limited patience for tech ownership, as might consultants friendly with them be flat out counting past 10 on reasons why getting tech-dirty are smart, or even essential to surviving/winning inevitable industry competition and consolidation. It's human nature to stick the the sauce we know and believe we can only reliably cost as much, particularly so in incumbent corporates answering to boards preferring their projected returns low-passed for predictability and familiarity.

CSMS love plus costs

So with that, let's be honest: managing charging infrastructure is hard. Not rocket-surgery-difficult, but there's some decently technical stuff to wade into.

Accordingly many CPOs love a software vendor that appears able to solve all the nerd bits of a problem with a high or otherwise abstract technology payload from cloud sold as SaaS: "right, so we pay these people, no one needs to do anything physical at site, that's the tech dealt with, now we go make money" - nothing unfamiliar here.

This explains a typical segregation of CPO and CSMS businesses, and why some of the latter that held CPO assets historically divested themselves of as much. So right now (for the most part) we have CPOs and we have CSMS vendors that work with them.

And as the technical evolution of vehicle charging goes deeper down a complexity rabbit-hole - essentially trending towards a Tesla-like experience and then some, with the (very much not) small matter of every blue-dot-to-blue-dot stage being intended completely interoperable - CSMS vendors are either directly or adjacently able to solve more problems, if at a cost:

• OCPP Security Profile support or OCPP 2.x? That might cost you money.
• Load balancing? That might cost you money and depending how you want to implement it, it will likely cost you money.
• Support for eRoaming? That'll cost you money.
• Plug&Charge? That'll cost you money.
• DSO/smart grid integration? That'll cost you money.

That's the 'yep, I very clearly understand it's a part of EV charging' stuff - but many are developing capabilities to run adjacencies too:

• Want to bid aggregates of EVs as part or whole of VPPs? In energy and/or services markets? We can do that too, and... that'll cost you money.
• Want to control batteries at site to manage EV charging demand? See above...

...I could go on - suffice to suggest that intrinsic CSMS costs on a per-plug basis aren't trending towards zero. Many readers will identify some of the latter opportunities as So Hot Right Now; whether you use it or not there's a lot of capability in the above that needs to be maintained, a lot of risk that needs to be managed, and even with scale the amount of value ceded to CSMS vendors in future will remain sizeable through second and third horizons. To be fair Australian demand for the more-esoteric stuff exceeds that of the markets that typically supply these solutions, so it's very roadmapp-y if credible for PE firms and other investors much liking the sound of that cash register leading to observations per the below:

Teddy Flatau's observations aren't surprising: if you're in the low-power charging game and not based in China your prospects may be looking somewhat niche in market ('niche' is not 'dead', simply to suggest that expectations around owning the majority of a commodity market in lieu of cost leadership are not generally realisable) + a quick scan on Crunchbase will tell you that CSMS vendor valuations have indeed been on a bit of a tear or late.

A Reality Helmet for that CSMS love

Those valuations are broadly based on the simple premise that the world will have tens to hundreds of millions (or more) charging stations that will need to be managed; that the need for management will extend beyond public or otherwise shared charging infrastructure and into the residential sector; that organisations that wish to own or monetise products or services from this infrastructure will need to pay rent to firms that will make them dance as they need for any requirements from market compliance to revenue generation.

But there's reason to pump the brakes on this logic too: no sane, capable procurement executive anywhere in the world that would advise any organisation to "build a business with billions-of-revenue potential that is completely contingent on one third-party vendor needing deep and difficult-to-transition-from linkages into your business to work, which has a likely variable cost base as the industry works itself out as to what it needs to do functionally, which you don't have a controlling stake over, which forms part of a rapidly-scaling-and-evolving industry and which may be acquired by a competitor at any stage."

There's a point at which relevant execs look to the future and realise:

• CPO growth (often, not always) isn't in the volume minority of DC-fast infrastructure, it's (increasingly) in higher-volume, lower-power stuff,
• From CPO perspectives these costs don't scale - if ambitions are hundreds of thousands of customers with >AUD$100/year liability to a third party to run their EVSEs (or more) - will your business actually hit that scale with those intrinsic costs if a competitor finds a way to do better? (Anyone in retail energy will tell you - we're talking a 'customers will shop around' amount of money right there),
• Whilst much of the CSMS cost stack is recovered across all customer assets, hardly any of the customer asset fleet needs everything included in the CSMS cost stack - different charging station scenarios will often need different feature sets - and there's often a bunch of stuff you're paying for that's there because a CSMS vendor needs it, not the customer (e.g., well-maintained hooks for CPO/eMSP billing platforms - any given CPO/eMSP will have one, not a need to support an industry of as much) - this is particularly relevant when many CSMS vendors are based overseas and develop against majority interests representing market contexts foreign to our own,
• Conversely, a situation in which (your) CPO business is limited in what competitive solutions you can bring or otherwise develop for your market by a key solution vendor it pays (a lot) of service money to with limited sway may become untenable...

...It goes on. Usually at this point someone starts to embrace the notion that the odds of a third-party CSMS vendor competitively, strategically rocking your corporate world towards large-scale industry leadership and pots of gold at the end of every rainbow... ultimately asymptote to zero. That long-term, go-big-or-go-home success in the space really does mean that you need to own your own CSMS, that it needs to be built for your needs only, and that the lessons learned in doing as much - if correctly and proactively absorbed in your business - go a long way to creating the appropriate relevant-technology-aware culture any business serious about the space needs to foster in order to lead it.

Other reasons?

Some other reasons to care (there are more):

Care of charging infrastructure is a tough gig

OCPP will talk to your CSMS in general, interoperable ways. There's certainly scope for customised error messaging if you're snooping every OCPP message every EVSE spits out though generally your interactions with a CSMS will be architechted towards APIs with higher-level stuff.

In practice there's a lot that can and could go wrong with EVSEs in the field, and tighter integration with relevant data streams have it over generalised approaches for leading asset management - if you know something's wrong and are about to send resources to site, knowing what they're likely to need to do towards success can make very significant differences to OPEX. So can data-driven understandings around predictive failure.

Getting it right can mean very significant differences to your top line and growth - ask your averaged EV driver why they charge where they do - most of the time they'll answer 'reliability'.

(Beyond Tesla) this is a very poorly-served area of charging infrastructure management. Building relevant applications on top of third-party CSMS APIs is possible, but fiddly and limited when your business ultimately rolls CSMS vendors. But build it yourself and you can do a number of things - from more efficient application design to going beyond OCPP to leverage whatever service protocols the EVSEs you've chosen may support.

EVs as DER really isn't trivial

Most CSMS vendors will claim they can load manage into the nth dimension, that flexible trading as DER is their or their roadmap's thing, the PFR is coming and that stationary battery control over OCPP is coming. They'll have an integration with with a physical metering platform provider (or a few) to be able to follow variable net limits appropriately.

All true but neither really competitive: consider that Australia is the leading market for DER proliferation and that by virtue of the volume of kit in the market and the experience with energy and services market designs, we have a more mature idea of how to do this and why than most other markets:

• We've been there with platforms that only manage one DER and are passive to the rest - it often doesn't scale to meet customer needs, which may evolve to multi-DER and/or high-value (if sometimes complex) site integrations,
• The value obtained in flexible trading implementations by physical control hardware at the edge - in lieu of regulatory distortion - exceeds that obtained by cloud-only implementations, and can handily recoup the cost of enabling hardware,
• IEEE 2030.5 allows one endpoint per premises from which DSO directives flow - doing it from a CSMS just for the cars won't work, and
• And above all, why cede control of DER performance - particularly where flexible trading's supposed to happen - to third parties affording limited transferability of performance across vendors and who have limited (if any) regulatory/policy representation in your local market? This is a space evolving practically by the day; if efforts aren't present, how can they be expected to have the currency needed to give market-competitive outcomes?

Forget the top line estimates of EV as DER value (particularly V2G and FCAS) - most are wrong/heavily inflated - and start realising there's many costs to trip up on here en route to creating value which may (in whole or part) be transferred to customers. EVs as DER will have significant volume potential but their lower intrinsic availability (as tradable elements) shifts our focus away from what the revenues could be and more to what costs we could maximally sustain; the companies doing this successfully in Australia - which can be counted on one hand - have spent years bringing relevant capabilities to market as core business.

It's accordingly no surprise that world-class solution providers for EVs-as-DER specifically have somewhat flamed out, or that the vendors currently doing EVs-as-DER load management and flexible trading best started life as DERM vendors and branched out - if you need one, get one, or dedicate time to lowering access barriers through proactive reg/pol work, or both. You're just unlikely to find winning success through an add-on to a SaaS offering from an adjacent industry provided by a firm with a native market of 'elsewhere' and limited relevant, proactive investments 'here'.

Security can be underbaked

Original-flavour OCPP runs a fairly standard username/password authentication, it's up to you if you encrypt the channel, no server authentication. Obviously, this ('Security Profile 1') isn't best practice.

Some vendors will talk to cyber teams and try to run the lot in a tunnel, however OCPP has Security Profiles that are more expansive (2 and 3). Both encrypt the channel with TLS 1.2 or greater, there's a server certificate and client authentication can either be password (2) or client-side certificate (3). Despite not being new they're often not used, however. Many CSMS simply don't yet support them.

Without going into a (very) long article it's enough to suggest that cyber types will point out that (a) the better profiles should absolutely be used and that (b) they may go a long way but are not a complete solution to cybersecurity.

Bring it in house, however, and can leverage the best en route to building it how you want.

Sometimes it's not what this costs but where

On the one hand some features of a CSMS value stack offer marginal value in some scenarios - strategically it's worth considering whether those costs are better absorbed elsewhere. Some examples from the above?

If a business is concerned with getting grid-connected EVs to respect DSO commands it's fair that a utility would this capability across its entire DER asset fleet, not solely EVs. This gives rise to a spectrum of more competitive cost stacks, from bundling HEMS with EVSEs to Octopus bundling Kraken + KrackenFlex (Origin Energy's looking quite smart with 23% ownership there) - a dollar spent in these approaches can go a lot further (given the ultimate intente of many businesses in this space) than spent solely in a (SaaS) CSMS offering; similarly it's reasonable to expect that over time and given a greater income base, such approaches may be the recipients of greater development resources towards becoming investments that are better under any business conditions.

Cybersecurity approaches are another example - whilst executions can differ across DER types (particularly where certificate-based charging session cost recovery is concerned) most affected businesses will face similar challenges with other DER types, and will strategically look to not invest twice for similar outcomes to whatever degree possible.

Costs and ambitions

There are give-or-take 10M cars in Australia. If half an EVSE assignable to them (a finger-in-the-air low bound) that'd be 5M EVSEs. If they each had $100/year to a CSMS for management (again, low bound) and a business targeted 100k cars, that's:

• 2% of market (all the major CPOs here are hoping to grow to more than this!)
• $10M to a CSMS vendor a year (ex fat in integration efforts for application development, ex the value of many other lost opportunities in as much)

Whilst the above numbers are coarse and whilst CSMS firms might be undergoing a hype cycle with valuations at USD$150-400M, it doesn't take anywhere near that to build a workable CSMS. We know because we've done this a few times in Australia (there's even been an acquisition; Australian Motoring Services is looking pretty sharp on that right about now) yet the space has matured considerably since and all sorts of dev resources now exist (from this to things far more commercially-oriented) towards competitive results with less effort (that's not to diminish ChargeFox's efforts - modern Australian EV uptake owes ChargeFox a great deal).

If serious about sticking around for scale volumes what's needed to stand something up gets by on NPV pretty quickly. Yes, much of a Board will intone that scale won't come for some time but would be similarly be remiss to not appreciate that it's (a) inevitable and (b) that the CSMS cost realities involved only become more of a handbrake on scaling competitively.

There's some cost optimisations to be had too; we're talking about large-intended CPOs standing up a CSMS to serve themselves only - needing to only support the EVSEs as used and building a solution capable of integrating with one eMSP are two obvious examples of this, but far from the only ones.

Go shopping, or...

It may be worth mentioning that Ampeco - which currently has solid revenues and which white-labels CSMS offerings to very many vendors globally - Series A'd at just $16M last year (which'd put its total valuation within striking distance of many large corporates in or eyeing the Australian market - most CSMS firms reaching Series A raised similar money) and is only five years old. For Highly Corporate Ears 'only five years' might impart an impression that we're talking to a relatively immature offering - not so - Ampeco has a world-class product. They're not the only ones with similar starts and progress; Wevo Energy was acquired by SolarEdge recently having a very decent product in market... and having started three years prior.

This article points to a market for a not-quite-Ampeco which might be worth less. In lieu of one, strongly consider building from scratch.

This doesn't require a complete leap into unfamiliarity:

• There are, evidently, people in Australia with direct and successful experience in standing up a CSMS,
• There are also many other businesses that can be acquired or otherwise invested in with adjacent in-domain skills to get you there. Some of these businesses already do relevant things that aspirational mega-in-market CPOs need to be successful, be it distribution, installation, aftersales support, EVSE build/supply, etc. All are valid perspectives from which to contribute.

The point is that (as ever) it takes good and relevant people, and even if they didn't specifically lead CSMS development for their last job there's talent in our shores (or beyond) with EV-industry knowledge that certainly can.

And if it's any consolation towards the art of the possible, CPO and related business archetypes that focus on a specific segment rather than the entire mass market (e.g., apartment/strata charging, truck charging, etc) are - by virtue of trying to build competitive, specific offerings - quite often building CSMS platforms of their own. And they work just fine.

Moreover it's not difficult to imagine that organisations embracing - not deferring - competence in underlying technology as core may face the technology challenges of this rapidly-evolving, technology-enabled industry more successfully than organisations that don't. It worked for Tesla and a range of other firms; whilst it's not everything required for success it certainly goes a long way, to which ends there's real, tangible business value in doing so.

Final note - Real Competition is yet to reveal itself, but it's out there

This article points to a local return to integrated CPO+eMSP businesses that roll and and run their own CSMS.

These companies do exist overseas - they'll come here eventually. Some are brands we've known for years (e.g., Shell didn't pickup Greenlots by chance). Some won't be familiar to most of us, and they're the ones worth looking at in serious detail; take a look at this article. It's not every vendor mentioned (nor is the list complete), however if you're a company that:

• Makes it's own charging infrastructure in a country with leading, low manufacturing costs,
• Has a CSMS that works with your product, along with other indigenous resources to plan and deploy network infrastructure,
• Builds equipment to the interoperable standards mentioned earlier, and
• Has valid experience running a CPO business and eMSP business in addition to integrating 'more blue dots' per the start of this article,

It's likely your impact in a market like Australia - where EV users are just getting into reliability and cost to serve sensitivity - could be very, very profound. About the sleepiest assumption corporate leaders/advisers here could make is that Chinese vendors will continue to sell widgets only and avoid getting into service delivery. They'll come, with highly-developed integrated offerings having a cost base optimised in ways we're yet to experience as a market. Some already have integrated offerings that are objectively - from a range of perspectives - superb.

Of most of the CPO and adjacent businesses I've met locally, only a minority has the above considerations in their strategic planning...

...Makes a case for a strategic sharpening of pencils, then.