Murthy Nukala is CEO of Adchemy, an online advertising technology company specializing in search engine marketing tools for advertisers.
As part of its "State of..." series of articles with industry executives, AdExchanger.com sat down with Nukala to discuss his company, his views on the space, and the state of Adchemy today.
AdExchanger.com: What would you say is the high-level update on Adchemy's positioning?
MN: So the highest level is that we think there is a lot of innovation and opportunity left in search, which people consider to be primarily dominated by Google, obviously. It's a $15 billion market in the US and $25 billion, globally. We think that the market is this big, almost in spite of itself because in spite of the fact it is grown as much as it has, it's still fundamentally very complex and very inefficient for marketers.
The complexity stems from the fact that the core, underlying unit is still the keyword. Keywords are very problematic for advertisers. First, all keywords have a massive tail and there are many different ways in which consumers express their intent. The way you and I express our intent in keywords could be vastly different even though we have the same underlying intent.
And, advertisers have a very hard time converting products into keywords. What they know, understand and sell are products. But taking each product and figuring out out how they convert them into keywords - that tends to be very problematic – especially if you are selling tens of thousands or, in some cases, a 1,000,000 products or more. It is close to an impossible problem.
It's a problem for creating relevance, too. If you have a small number of keywords that you are advertising or in your portfolio, then you can actually achieve relevance. But if you have a large number, you typically have a very broad coverage that tends to be very irrelevant.
So those two things - keywords and relevance at scale - led us to invent something new that we call an equal interpretive extraction layer - or "Intent Maps" which essentially simplify and scale an advertiser's campaign. So you can get the exposure to a dramatically larger number of keywords. But with the same amount of effort that you would expend against a much smaller set. Because you are dealing with roughly a thousand intents, as we call it, to manage millions of keywords.
There's an order of magnitude reduction in complexity and at the same time, if you're simplifying scale, you also increase relevance. It just has good things all around for advertisers on the one hand, but also the publisher and for consumers.
This product used to be called "Word Map," right? Why the name change?
You are right. And the reason is because we realized that intent goes beyond keywords. We started out saying that Word Map is really about keywords and then we realized that there's a big shift towards what we call non‑query intent. We think of there being search by query - which is how we know it today. But also search by behavior [through the use of] mobile apps and things of that nature.
So these are substitutes in a way and we realize the intent actually goes beyond keywords and so we call it the intent web, the non‑search web. And therefore, we call this Intent Map, not Word Map.
Google announced recently that it's going to offer the opportunity for advertisers to have a plus one button inside their display ad unit. What are your thoughts on that and secondly, how can Adchemy tap into this sort of information, whether directly or indirectly?
We think that's actually more relevant for [understanding] the connection between users and ads. It's a way to capture more user‑centric ad data. We actually stay away from users as an organizing unit. We think of intent as the organizing unit.
So in many ways, it's interesting… it's basically in response to Facebook in my view. Where Facebook has a lot of connections between likes of the user and ads. And this is a way to replicate that. But we have a different angle – we say groups of users that have the same intent should be treated alike and we don't either track or deal with users on a user by user basis.
So for us, that's a very important design principle.
Regarding Microsoft's investment, what is going on there generally speaking? It seems like they're taking a more active role in trying to shape the advertising ecosystem. Do you agree?
I think that's a fair statement. I think that they would actually say that that's not just a fair statement but an explicit strategy - taking an ecosystem approach. I think it's both healthy and enlightened as an approach because clearly a lot of innovation happens in the ecosystem. They're saying look we can both tap into, as well as encourage innovation, and as in [Adchemy's] case, invest in it.
How important is Microsoft's adCenter for search marketers today?
I have two responses or reactions to that. The first is that we have seen in the market that there is a genuine desire to increase spend on adCenter. There's no question about that.
Is it because they are capped out on Google?
I don't know that they are capped out. I think there is a genuine interest for a number of reasons. As a marketer you don't want to be concentrated. You want to make sure that your traffic is coming from many sources. And that's why almost everyone we talk to wants to increase their adCenter spend.
The second thing is that I would say is that [in regards to Microsoft's invenstment], this is a non‑exclusive partnership. And equally applies across any advertiser's Google spend as well.
What this really is an extraction layer across all search. And basically the core message is that there is a more powerful way to manage search - by managing intent not keywords. It's such a fundamental concept that we see people using it across their entire search string.
Going back to the investor group, I believe your previous round of investment included Accenture. I notice they're not part of this round. Where is Accenture fitting in your plans today?
I think we're a very lucky company in having two committed partners who believe in us - and who also work very well with each other. So in some ways it's just a fortunate place to be. We have several very successful engagements together with Accenture that we are not announcing yet. We just see the Microsoft partnership as a very exciting extension.
Back to the product, any thoughts about rolling out a product that's targeting the display market?
The core focus today is search. What we actually see, however, is that intent is dispersed. We see consumers expressing their intent in a lot of ways.
Mobile applications are an important target area. When I think about how we define our market, it's what we are calling the intent web, which is that part of the web where intent is expressed and can be detected. There are pockets of display where that's true. There's obviously a large part of mobile apps that actually fall into that category.
Search is creating the strongest signal. But we see ourselves extending into other channels. That's also one of the reasons we call it Intent Maps and not Word Maps because we feel like Word Maps ties you very closely to search. Whereas, Intent Maps goes across search into other parts of the web where intent can be detected.
An industry-wide question for you - what is your overall take on the ecosystem today? Are there too many companies in this ad tech space?
My view is that there are too many undifferentiated companies in the space. That is the problem. There are too many people doing the same thing. When you have that happen, you essentially end up with a lot of supply and very limited pricing power – and that is bad for companies.
There were several markets, which went from introduction to commoditization in months. Usually you have an umbrella of a couple of years. That is because, I think, when something new is introduced there are a lot of companies that start calling themselves [by that name] and essentially flock to it.
I think there is going to be a shakeout in the ad tech universe. I don't think there is any question in my mind about that. It's actually not a bad thing. Technology industries always goes through cycles. This is a part of the cycle, where you want the hottest part to burn. You want it to clear out the brush. You want the strong to survive.
With search we are talking about being part of a just‑in‑time marketplace, non-guaranteed in a way. How do you see a product like Intent Map fitting into guaranteed markets, either now or down the road?
My view is that it's less about guaranteed than non-guaranteed. That's really as much of a sales question for publishers as it is an inventory question. The reason you have guaranteed or nonguaranteed is that you have two vectors. The first vector is – "Is there a way I can determine who my most valuable versus less valuable users are?" Typically users you have that have data associated with them tend to be more valuable than those that don't.
The second vector is, "Can my sales team actually sell this in a guaranteed way or not?"
The vector that I see us actually tapping into is the first one. The more you can understand about the intent of the user within any given context of your site - or however you choose to engage the user as a publisher - the more you a have set of users that you can actually sell guaranteed inventory against.
So, in some ways we would increase the supply of valuable inventory for publishers. What we wouldn't, in effect, do is the second dimension, which is really how your sales team goes out and sells against that.
Specifically, our focus would be primarily with mobile app developers. Because clearly the mobile space today is in the same stage that display was in the late '90s, which is completely undifferentiated. There is no notion of relevance anywhere. It is a wild, wild West in terms of relevance to the user. It just doesn't exist.
We actually think that IntentMap plays a very important role in getting a high level of relevance to the mobile app publisher. Whether they are able to sell that is guaranteed or relevant is up to the quality of their sales teams.
Is there a data opportunity for Adchemy as in re-selling user intent data?
I'm glad you asked the question. We absolutely do not track users. It is pretty important to us that that is very clear. That is not the reason we are in business. The reason we are in business is because IntentMap is fundamentally a more powerful abstraction layer for search. That is really the goal, to be the focus of everything.
What are a few milestones that you would like to achieve a year from now?
I think all of our milestones, as well as the reason for the funding, are all very aligned.
We have shown great early results with IntentMap. We are looking to have Intent Maps be ubiquitous. We think there is a lot of work and opportunity in taking this into the broader advertiser ecosystem. That is the focus of this [Microsoft] partnership. It is also the focus of any milestones that we have. It's all about penetration and adoption.
Is Adchemy ultimately trying to put together the proverbial "marketing stack"? Does Adchemy need to provide an end to end solution at some point - and across channels?
I think your question actually has very strong display roots. It makes a lot of sense in display. The reason for that is essentially because the strategy in display is about a unified cookie pool. It's about getting everyone on the same cookie space. That is basically what DoubleClick is accomplishing with great success. Whether it is the publisher or the advertiser or the exchange, the more that all of them are in the same cookie pool, the more that you end up with economies of scale. Therefore, in the display advertising ecosystem, because it is so cookie‑centric, you need the unified stack.
Now, there are a couple of things that are happening in ecosystem that essentially make that less important. The first thing is essentially that consumers are going across multiple devices. Where you previously had just a laptop or a computer you now have a tablet and a phone in addition to that.
Because of this device proliferation, we actually think cookies become less relevant and less important. It becomes very difficult, because there is no such thing as a multi‑device cookie, except for signed in users and there will only be a small fraction of users at any given point in time that are signed in users.
So we think that, because of device proliferation, cookies as the unification layer become very problematic. It becomes very difficult. So I think that in that world, where you have device proliferation, the traditional thinking about the marketing stack, which itself was driven because you need a unified cookie space, actually is less relevant in the future.
What is relevant in the future is what your new organizing principle is going to be. We actually think that that organizing principle of the future is intent because it's very privacy safe, just like search. You don't know who the user is. What you are reaching, if you are building on a keyword is a group of users who already typed in the same query, except you don't know who the user is.
It's very similar when you use intent as the unification layer across devices, because you don't really have to care about who the down the line user is. Most of the time you can't. What you care about is what their intent is.
That's why we actually think that our role in the whole ecosystem is to be the platform for intent and to be the way by which, at any given point in time, no matter where that user is, you actually have a way for a publisher or an advertiser to have a structured way by which to understand intent and then to also leverage it for marketing.
So that's really the role that we play, which we think actually becomes more important as users actually have device proliferation.
By John Ebbert