The Rise Of The Mega Stacks

Danny Hopwood, Vivkai, head of platform, EMEA, gives his assessment of the current programmatic landscape in light of recent industry consolidation.

At the risk of overusing the latest buzzword, we are in a period of serious consolidation.  It was inevitable due to the impressive explosion of new technology companies and also the many pivots made by the old guard into the programmatic space.

The effects of consolidation are essentially just different forms of pressure.  Pressure to perform on a plan, pressure to change your business, pressure to build to be bought, pressure to merge, pressure to educate, pressure from your VC investment relationships.

It is this pressure that is forcing decisions at every major company within the programmatic space.

So let’s start with the consolidation effect, specifically around digital ecosystems.  Currently we have five-to-six big digital ecosystems forming.

1. Google has accomplished impressive integration across its platforms (DBM, DS3, GA, YouTube etc), analytics and the ad serving business, not to mention its data tools like big query.  What this means is that with a few clicks and authorisations I can advertise to the same consumer irrespective of channel.  There are still a few areas for improvement, specifically in mobile and video, but on the whole you can reach a consumer across all digital channels.
2. AOL’s recent announcement that it is going 100% programmatic was maybe to some a strange move but to most in the programmatic industry a welcome change.  AOL is aiming for the same result as Google.  An ecosystem where you can see a user through its technology across multiple channels.  AOL has a strong video proposition with the Adap.tv acquisition last year but its mobile area is either going to have be built out, bought or borrowed.
3. Amazon launched its self-serve offering last month (disclaimer: VivaKi is the first agency trading desk to run the tool).  Again, Amazon is following the same route – one big Amazon ecosystem to reach consumers across channels programmatically.  Its recent phone launch shows Amazon is thinking about not just the purchasing behaviours of consumers but also the connection of that user’s desktop profile to their mobile’s OS.
4. Yahoo! launched its ad manager+ platform and revamped RMX to remove those pesky publisher integration contracts. Criticised by many, but it is genuinely trying to put the building blocks in place to develop its own ecosystem.
5. Microsoft is an ecosystem but currently looks inconsistent with those listed above in that the majority of the programmatic activity is run on technology it does not own.  Microsoft has the characters and attributes of its contemporaries but is missing the owned technology element.  Its investment into AppNexus binds it for now to the AppNexus tech stack.  And here I think a special mention has to be given to AppNexus for a solid open platform that has locked itself into an ecosystem. The problem for Microsoft is that it is only as fast as AppNexus in programmatic. This was evidenced in private marketplaces only being made available widely in 2014, almost two years late.  Microsoft can’t be faster unless it looks to someone else to power it – or it makes an acquisition. Furthermore, I just get the feeling this is still not on the radar of some of the senior people at Microsoft, which may be down to a company that suffers too much red tape, or makes enough money in other areas of its business to worry too much.
6. Facebook has long been called an ecosystem and is slowly but surely beginning to show signs that they might reach the same position as Google and do it quicker than anyone else.  The Atlas acquisition was smart and despite many saying the adserving business would suffer, there are still plenty of Atlas users out there.  The next logical step would be to build a DSP.  Speculation for now but you can be sure Facebook won’t miss the opportunity like its contemporaries in locking in spend via inventory, data and tech that you can only use via a Facebook DSP.
7. Rubicon and companies that look like Dunn Humby/Sociomantic are the other players in this space. Rubicon needs to get work on the integration of data from all the premium publishers it has residing within its supply that could then lead to forecasting and first party overlay of advertisers data to determine endemic, and non-endemic supply for advertisers. I think Rubicon could be in a good position to be the ‘publishers choice’ if the industry forms into these big five-to-six systems. Dunn Humby/Sociomantic on the other hand has the data, the execution layer done and dusted.

There are others who didn’t make this list because it’s still unclear what their moves are.  Twitter with its MoPub move certainly shows it might be aiming for the same goal but it’s too early to tell.

What’s stopping others? 

Once you start to broaden the scope, this very question arises.

Sky is in this bracket and is slowly but surely working on its offering.  Other broadcasters could fall into it. But I don’t see the desire from it past order automation for TV spot buying.

Telecoms providers also fall in here but it may be too little too late and the market may move too quickly for them to get everything lined up in terms of tech.

What unites the seven companies listed above is their unique elements across tech, data and inventory.  I believe if you have these elements, you own them and you’re BIG, then you have a strong case for an ecosystem.

Publishers’ options?

For those that don’t (you can put local market publishers, newspapers and standalone websites into this bracket), there are two options.  Join an ecosystem or try to survive outside of it (where Rubicon and AppNexus could help the market).  This is really the same choice as becoming programmatically enabled i.e. will you trade programmatically or will you ignore it.  I think we can confidently say the market has made that decision for them although the conversation still rears its head of not cannibalising direct spend and will continue to do so for the foreseeable future.

And what of the VC funded DSP and technology companies?  

How and where do they fit in when their tech becomes a commodity?  I mentioned in my previous post that you can live without certain features if you have other basic features while you wait for your technology vendor to build it (or you build it yourself).

But the commodity aspect is the buying.  Execution in its simplest form on all the DSP’s is the same to varying levels of success/performance.  What differentiates a DSP is what the ecosystems have that is theirs.

I don’t need to buy a standard third-party data segment in a standalone DSP if my ecosystem DSP has that third-party data segment plus all the unique elements that they have, and no else has it in their DSP.  This is the crux of it.  If buying is all you deliver, how can you differentiate yourself?  Service and product can all be replicated.  You have to have something no else has if you’re going to succeed.

Or you get bought by someone who needs that buying function.  Microsoft, Yahoo (post-Alibaba) could answer that call – if they want to.  I also expect these VC funded DSP’s to begin making a play for the marketing cloud landscape. Rightly or wrongly that’s the next big bet for these companies.

It’s the same situation for any channel specific DSP in mobile or video.  We have talked for years about a multi-screen approach, cross channel planning and execution but the mobile DSP caters for mobile, a video DSP for video.

The convergence of channels in DSP’s is happening and it is the big ecosystem DSP’s that are going win out, especially if they own their video or mobile inventory.  To a degree the topic of cross screen statistical modelling is redundant if you own the three pillars of tech, data and inventory – you can connect it anyway within your own ecosystem (I make this sound simple – it’s not).  But it will be valuable for those that don’t, and if a standalone DSP buys a cross screen solution that’s further consolidation.

You can be absolutely certain that these ecosystems will not for a minute forget about mobile or video.  When they turn their focus to it, they will build, buy or borrow. A thought - maybe that’s why we have so many standalone channel specific DSP’s – build to be bought!  Ultimately they want what most clients want – one single view across all channels in one platform.

Interestingly though the ‘one view’ won’t be one view.  It will be five-to-six dependent on how many ecosystems you trade within.  So the next logical step is to try and comprehend how you track across five-to-six ecosystems that have unique data, tech and inventory across all channels but more importantly have a different ID for the consumer.  A Google ID, an Amazon ID, a Facebook ID and so on.

This requires the big ecosystems to play nicely and somehow allow a cross ecosystem tracking option and also a technology company to be that cross ecosystem tracking solution.

Maybe this is a DMP, maybe it’s something that doesn’t exist but it won’t be a Facebook, Google or Amazon.  They’ve spent the best part of 10 years and billions of dollars in tech to lock you into their ecosystem so it’s unlikely they will just give you an option to trade correctly with their competitors.

The prospects of market neutrality?

I think this has to be someone else, someone neutral but with enough scale and clout to get support from big advertisers to ensure this could become a reality. This ideally would be an Oracle BlueKai or Adobe, someone who occupies that cloud based marketing platform arena in conjunction with the advertisers or agencies. This goes beyond just the DMP to adserving platforms and buying platforms, but overall agencies are in nice sweet spot of convergence here.

The decision on execution platforms be it standalone DSP or ecosystem DSP’s best lends itself to the current agency trading desks and skillsets that sit within these areas.

There is a reason we built VivaKi Verified and it’s not just to verify the tech, data or inventory we buy but to understand what DSP is to be used where.  Initially there wasn’t the potential for multiple ecosystems but now this becomes more a reality.

The requirement to understand the benefits of every piece of the ad tech industry becomes more important to effectively communicate back to clients why this DSP, or this ecosystem becomes their primary arena for trading.  Potentially with secondary and tertiary ecosystems or DSP’s to answer other needs.

What’s important to realise here is that no one ecosystem is going to be able to do everything so if you’re using one ecosystem DSP, or one standalone DSP you’re going to be hindered in your ability to provide access to the other ecosystems. This is a new landscape, prior to previous 12 months Google was the only ecosystem.

This also means for the client direct angle for a DSP it could only ever be a short term prospect if the future becomes an ecosystem future.  They will only ever know their tech and its capabilities. A standalone DSP isn’t going to be given Amazon data, or Google, or Facebook data, or their prized inventory.

Plus I’m sure a standalone DSP wouldn’t want to have skill up on another DSP technology.

Again, agencies are in an advantageous position to give that intelligence and recommendations on initially how to trade in this space and who to work with. But it may need a change in DNA for some?

Is consolidation a bad thing?

Some will comment on the fact this consolidation is a bad thing; that the ecosystem future is terrible for the industry.  I disagree.

It cleans the market up it actually allows you to understand what your value proposition is. It has never been so clear that if ecosystems come into place for you to provide value to a client it has to be something they can’t get anywhere else.

This logic didn’t affect the broadcasters negatively and in fact got them where they are today.

Look at how well TV has propagated with a selection of five-to-six big channels that you trade across.

The difference in programmatic is that the act of buying and decision is in the hands of the advertiser or agency/independent that represents that client.  Not in the broadcasters once the order form has been shipped from the agency.  This is likely one of the reasons we see slow movement from the broadcasters, this and the bigger reason that they don’t need to right now.

Closed ecosystems, good or bad are inevitable

You have to figure out how you’re going to work in them, operate your business and most important provide inventory, data and technology that is unique and cannot be replicated.

Right now this is my perspective on what could happen, the building blocks are falling into place across the largest and most prolific digital bodies that we interact with daily, and for now we can’t rule this out.

Programmatic trading is here, some companies will be around for the future, some might not, and some that might not exist right now might be the glue that finally provides that ‘one view’ to rule them all.