springload

(Originally published in iMediaConnection, September 2011) by Eric Picard

The other day I was talking to a good friend of mine who is on the executive team at a startup in the ad technology space. We were talking about strategy — and in the midst of the conversation, I suggested that since the company hadn’t taken any money yet, it should strongly consider selling at its early “life stage” for $10-20 million now. He gasped and told me, “Are you kidding me? I’d put our valuation at between $100-200 million.”

I stopped talking for a minute, and then said, “I’m not sure how you could possibly have the revenue to justify that.” This is, after all, an early stage startup that only has been in business for a year or two, and I’m fairly familiar with the company and its customers; I know it’s not doing more than $2-3 million of annual revenue right now.

And he said, “In (insert niche here) nobody is bought on a multiple of revenue — it’s always based on strategic value.”

To which I replied, “That might be true of funding — of course VCs and increasingly PE firms are betting on the long-term value of disruptive technologies. But for an acquisition, at least for a rational one, nobody is bought without some discussion of ‘comps’ for similar companies and revenue, and some ‘reasonable’ multiple is definitely a factor.”

My friend rattled off several examples of companies that have been acquired (for what I see as irrational amounts of money) lately and some of his beliefs on their revenue pictures — and we picked over the specifics of the acquisitions. And that’s when I started to get worried.

This conversation has been rattling around in my head for days now, and I have to say I’m concerned about the market. I see this space as primed for consolidation. The Luma Partners display ecosystem slide is a perfect example of much that is wrong in our space:

As dollars move between advertisers and publishers, the folks sitting in the middle are trying to find a way to strip off some money as it passes through the ecosystem. The only way they’re going to be able to strip some pennies off of the dollars as they flow through is if they provide some value back to the ecosystem. The problem is both the number of companies in this space and the exuberance of those companies for how they believe they’ll participate. Many are not realistic on what they should be paid.

There’s opportunity in this space; don’t get me wrong. I wouldn’t be invested so heavily in online advertising if I didn’t believe that there is a strong opportunity for me and my company. But let’s all be very clear about what that opportunity really looks like. The greater the provided value, the more money that the company in the middle can take away. So is the value a moderate improvement in efficiency — or a substantial change in value? How significant is the change? At the end of the day, the market will bear only so much being stripped away, so only those companies that have disruptive technologies are going to be able to extract significant amounts of money.

It might be useful to look at what percentage of spend various vendors are able to extract today. Let’s start with agencies, which are often the target of technology companies trying to find a place to disrupt the market through disintermediation. But that’s crap. First of all, the agency lives in thepower position in the ecosystem. And despite the kvetching of the technically minded who don’t “get” what agencies do (nor even the difference between a creative and media agency), agencies provide a lot of value to the advertiser (their customers). Agencies are not easily disintermediated — nobody has been able to disintermediate them so far.

Most startups vastly inflate the amount agencies get paid — typically the number that is thrown out is somewhere between 15 and 20 percent of spend, which would be freakin’ awesome if it were true. But those kinds of percentages went out of style in the ad space along with well-tailored suits, smoking a lot of cigarettes, and drinking whiskey and water like it’s going out of style. Most big agencies no longer negotiate their contracts with the marketing team as an advertiser; they negotiate with procurement offices and negotiate for fixed margins — very low margins, in many cases. They’d be psyched to claim 15 percent of spend. They’d be excited about 10 percent of spend — even 5 percent, in some cases, would be cause for ecstatic celebration.

OK, so agencies are not where the money pools. What about tech startups? The reality is that technology vendors take small percentages of the dollars out of the flow and make it up on margin and volume.

Ad serving is a great example of this. A third-party (buy side) ad server is typically getting between $0.07 and $0.15 CPM for its service. That is really not a huge amount of money. It typically comes in at less than 5 percent of spend — and at volume, and depending on price, it frequently is down below 1 percent.

In traditional media, typical vendors are well below 1 percent of spend as the money travels through their systems. But ad serving is commoditized, you might say (and I’d argue that before too long, most technologies are commoditized). Look at DSPs, which have been the much-laureled darlings of advertising technology for the last three years. There’s very little differentiation here. They’ve all commoditized out to varying degrees, competing only on price or service, or minor feature differences, rather than by disrupting each other. (And for the record, there’s nothing wrong with this — which is sort of my entire point.)

“But the DSPs are the future,” you might say. “They’re the ones who are bringing automation and efficiency to this space; they’re the future of advertising! Damn it!”

Well — yes and no. DSPs are playing in an emerging media — the real-time inventory market. In emerging media, the top-line media spend CPMs are generally higher. (Let’s not have any illusions here — it’s a product of supply and demand in which the amount of available inventory is low and the demand is high.) DSPs are in an emerging space where supply is vast, and demand is small (but growing), and they still are taking a proportionally large chunk of spend (8-20% depending on the contract and volume) because the market is emerging and the average deal size is still quite small.

In emerging spaces, the technology vendors typically take much bigger pieces of the pie. For example, look at ad serving back in 1998 — CPMs were closer to a dollar. Look at rich media vendors, which could easily pull close to $2 out of the ecosystem back in the early days. But the core CPMs of the media in an emerging market are higher. Look at mobile: In 2004, the average mobile CPM was between $60 and $80, and is now below $5 (depending on who you talk to). And when the CPMs are high, and the market is still figuring itself out, vendors can take a big piece of the pie. Even in paid search, which hasn’t seen the bottom drop out of CPMs (for very strong economically provable reasons), the percentage of sustainable media spend by vendors hasn’t been very high. The simple truth is that mature media markets are only willing to allow very small amounts of money to leach away between buyer and seller for “table stakes” technologies.

Does this mean that the online advertising space is not as “hot” as investors have believed for the last decade? I think this space is incredibly hot — and that there’s a huge amount of value to be created and we’re only at the beginning of it. But let’s be clear. Let’s look each other in the eye and not pretend that the dynamics of an emerging market are sustainable over the long term.

There are only two tricks to play out here: You either need to be the Donovan Data Systems of your market (i.e., you are indispensible, are taking a reasonable percentage of spend as the dollars flow through you, and you’re the stand-out leader in your space). Or you need to be the company that redefines the market completely (i.e., you will use technology to fundamentally change the way the market operates). And if technology is at the center of that disruption and technology is the driver of that fundamental change, then suddenly the rules are different.

What bothers me about the space we’re in right now is not only that it’s getting really crowded, but also that most of the parties playing in the middle are not adding the value that a full corporate entity needs to be adding in order to both create and extract the value needed. Most of these startups are really more of a feature rather than a whole business. But if they’re just a feature, what do they plug into?

The problem is that consolidation is not easy. It actually sucks majorly — for everyone involved. I speak from experience; I was on the deal teams for of a bunch of companies we acquired when I was at Microsoft. I was involved in the projects to consolidate those acquisitions, and I’m friends with a bunch of folks who were in similar roles at Google, AOL, Amazon, Yahoo, etc. And it’s just never easy. The buyer has this nasty problem of a new and generally incompatible technology, plus a completely different culture — both of which are super hard to converge successfully.

And what about when you’re getting bought? It only works out well for those who are fairly mercenary — the ones who ran after the idea because they wanted to exit well, and who were determined to exit well, and were plenty happy to exit as early as they could. But what about for those who are in love with their own startups, who see them as children? Great entrepreneurs I’ve met look upon an acquisition as an opportunity to get their struggling products the visibility and distribution might that they deserve. And it’s called an exit for a reason. When your company is acquired, it ceases to exist. It’s no longer your company; it belongs to someone else, who is very likely going to screw it up and kill it.

The trick for having a successful startup in this space and a successful exit (not only for the cash value, but to have your beloved business count for something going forward) is for folks to be realistic about both the value they bring to the table and the way they can be leveraged. And let’s not forget that in order to really be valuable when you are acquired, your technology has to somehow rationally live in the context of the acquiring party’s landscape — both technically and culturally.

Exit earlier rather than later if you can — while you still own a good chunk of the company. As a founder, would you rather have 30 percent of $20 million, or 5 percent of $80 million? I’ll give you some advice — earlier is better. Exit before you have to scale the thing up — before you have to invest in customer support or in operations, before hosting everything in the cloud stops scaling for you cost effectively and you have to invest seriously in capital expenses and need to raise a lot more money.

And please — build your technology in as abstracted and “ingestible” a way as possible. Please — I’m begging you!

But I digress. The reality is that there are a lot of companies that are stuck. They’ve taken a lot of money, but they aren’t the leader of their space or disrupting their space significantly. And most of them have become targets for new companies coming in and running after them — and either exactly copying them (further commoditizing them) or disrupting them.

It’s these second-generation companies that are the ones to watch. They’re typically bootstrapped and generally doing more interesting things than their established competitors. And they’re the ones who are most ripe for consolidation because they can afford to exit for much less money since they haven’t taken as much from investors.

The only question is this: What happens when they get acquired? And what happens to the middle of the market — those that have raised $15-40 million and that have stalled on growth and suddenly face a plethora of competitors? They had better find a way to get profitable real fast.

(Originally published in iMediaConnection, July 2011) by Eric Picard

Let me state the obvious — because clearly it’s not so obvious, especially to those of us working in online marketing. Most products and services are designed with a target market in mind. This market could be as broad as those of us with teeth, who hope to keep them healthy into old age, or as specific as 38-year-old women who want white teeth for their 20th high school reunions. The trend for the last few decades has been toward designing products for narrower and narrower markets — and using specific differentiation between target markets to drive sales and profits. And of course, with better targeting available all the time, the ability to hone the product to a specific subset of customers will become ever more possible.

The best companies use a combination of personas and scenarios to ensure that they are nailing the product requirements early in the design phase. These scenarios (sometimes also called use cases) are pushed into the hands of eagerly waiting marketers, who in turn get the product put into a strong series of marketing messages (and even the actual creative) that tie to specific target customers (the personas). The personas for which the product and marketing teams have developed their products ideally make up the basis of a media plan.

I’m frequently shocked at how few of the basics are used in the development of media plans for online marketing. And I’m frequently shocked at how products are released with marketing messages and targeting that don’t match. In many cases, the creative for online is either just completely different than the offline creative, or it has been so incredibly simplified for online that none of the powerful messaging from other media actually make it through.

So I thought I’d write a short primer for online marketers so that they understand the whole product life cycle and how they should be plugging into it. In advance, I’ll warn you that there are numerous methodologies here, and almost every company does this just a bit differently. So I’ll just push forward a simplified version of a typical process, and you should be able to apply the concepts as you stumble across them. And of course, if any companies you’re working with don’t use some variant of what I’m describing, you should be a bit concerned.

Product planning
In a perfect world — where there are plenty of resources, time, and money to properly plan a product — the model goes something like this:

Three to six months of market research are commissioned, funded, and executed to ensure understanding of the market demand for the product in question. This process begins with a series of ideas and invention, combined (at least for existing products and services) with feedback from existing customers, and is turned into a strategic plan for what product will be built.

In this process, the target personas for the customers to whom the product is designed to appeal are created. Ideally some market sizing is done to determine what the financial opportunity for all companies running after similar products and services might be — and what the specific opportunity for the product in question might be. Simultaneously, work typically is done to determine what scenarios will be supported in order to bring clarity to all members of the product team, from research to development to marketing to sales.

Example persona: Wealthy, highly educated, sophisticated urban empty nesters — Brad (64) and Sandra (62)

Example product: Online banking services for wealthy clients with multiple homes

Example scenario: Brad and Sandra live in New York City during the spring and fall, in Martha’s Vineyard during the summer, and in Killington, Vt., during the winter. They need a way to ensure that all their bills are paid on-time for all their properties, all year round, even when they are rarely there. This service creates a very clear portfolio of all their properties, and all their expenses, such that bills can be easily assigned to a property, tracked, and managed in a clear automated way.

Product planning is really the process of defining the opportunity at a broad level, and ultimately answers the question of why a product or service should be rolled out. The more discipline, time, and effort put into effective product planning, the easier the job of all the subsequent teams engaged in the process.

Product management
Once the product has been planned and approved, it’s time to build it. In this case, we’re talking about a software development project that will be rolled out via a website and a variety of apps across PC, phone, and tablets. The process entails defining the specific features, creating the project plan, working with the product development teams to ensure the correct product decisions are made, coordinating internal communications, and development of the appropriate key performance indicators (KPIs) to measure the product’s success in the market.

In most companies, the product management team is really the “product owner” and makes all the decisions and prioritizations of features of that product. Essentially, the team defines what will be built, leaving the how to product development. In some companies the what is shared between the product management and product development teams.

The key to strong product management is always being customer driven — which means creating very powerful and accurate personas and scenarios that always drive the “true north” of what is being built. This process should become the basis of what is handed off to the sales and marketing teams in order to drive the go-to-market strategy and sales positioning.

Product marketing
Product marketing is typically one of the most important teams, leading one of the most important efforts — but frequently this discipline is under-funded and under-resourced. In an appropriately resourced product marketing effort, key partners and customers are engaged in deep ongoing conversations. Ideally, the personas and scenarios that were created during the product planning effort received vast input from the product marketing teams. The go-to-market strategy for how that product will be launched, including all the marketing and training materials used by sales and customer services within the company, is managed by this team, which also feeds the key marketing positioning to the marketing communications teams.

If the product marketing team does its job correctly, the corporate marketing and sales efforts will be successful. Product marketing ultimately owns the decisions related to where and when the product will be rolled out (with huge dependencies on all the other teams).

Marketing communications
Given the intended audience reading this article, I won’t spend a lot of time here — as this is either you or your direct customer. However, a few key points are worth spending time on:

If the correct personas were created, the media strategy and even the core media plan should come together like a breeze. If the correct scenarios were chosen and executed against correctly by product management and product development, then the creative of the advertising should be quite easy to conceive and execute. In a perfect world, there is a direct feed from inception to creation to getting that product or service in front of prospective customers — and converting them to active customers.

There are, of course, many other teams involved any business, and all play critical roles at varying moments of the product lifecycle. Hopefully this rather nuts-and-bolts summary of the overall process will help those of you who have grown up attached to this mechanism either internally or externally, but who haven’t had full exposure to the processes and roles.

(Originally published in iMediaConnection, February 2011) by Eric Picard

Every mature media has one thing in common, and that is scale. Whether we discuss television, radio, newspapers, magazines, or out-of-home, they all have locked down their basic planning, buying, and selling processes in ways that enable a new employee in the space to learn the basics quickly, and everyone in those spaces has agreed on currency, methodology, and KPIs. Any two media planners in television can understand each other’s approach quickly, and can explain their goals to a sales person quickly, and can execute a media buy quickly. All with common knowledge within their industry — that is broadly available. This leads to scale — the ability of a marketer to reach large audiences in these media types at reasonably low costs per thousand impressions, and without a huge amount of work or cost to execute.

I’ve written before about the problems facing the online display industry, and how the early decisions made about ad serving technology are some of the drivers of the biggest problems we face. Essentially my belief is that because we took requirements from an emerging media — which are radically different from the requirements of a mature media — and locked them down at the heart of the inventory management systems behind the industry, we are screwed.

Emerging media have some common characteristics:

• Small amounts of available inventory, with relatively high demand, thus driving high prices
• Small overall budgets because they are coming from experimental media budgets, which are highly scrutinized and optimized during the life of the campaign
• Technical people are usually involved (i.e., experts with arcane knowledge of how to tweak the emerging media for maximum value extraction, across all phases of a deal, including sales, service, production, operations, and analysis)

Every emerging media type that I’ve touched, studied, and participated in over the last 15 years have all had these characteristics. From online display itself, to mobile, to in-game, to paid search, to rich media, to real-time bidding, I’ve seen this happen over and over. So why do I say we’re screwed in online display? Well, mostly for effect — to get your attention and see if we can dig our way out of the problem.

We built all of the original ad serving platforms, created all the processes for buying and selling inventory, set in place the KPIs, and invented ways of planning and measuring the effectiveness of the campaigns when the amount of available inventory was low, average deal size was quite small, and differentiation from other media was the driver of all the decisions. Not a bad thing in itself, but a horrible thing when we locked all those requirements down in software right off the bat. Because now it is nigh impossible to change the way those systems and processes function. And we really need to if we’re going to scale the industry.

I bring this up now because we’re going through the biggest revolution we’ve seen so far. Real-time buying and selling could solve all our problems. But the players in this space are falling into the same trap that all emerging media have fallen into, and if we’re not careful, we’ll have the same problems later that “standard” online display has today.

1. We need to reduce the amount of arcane knowledge needed to successfully execute on a real-time media buy. The market feels a lot like paid search in the early part of this industry, where only a small cadre of experts could really pull off anything interesting. Those people are all the ones leading paid search practices in the industry today. Good for them, bad for the space.
2. We need to optimize for efficiency over effectiveness. By this I mean that in an emerging media that is trying to prove itself, much of the effort is applied to a small frontier of effectiveness gains that will show numeric advantage over the competition. “We achieved 30 percent better results than competitors” sounds great until it is understood that the actual value created was miniscule. The big opportunity for the real-time space is scale, which is why I like the term “scale display” for this emerging space much better than anything else. Efficient (and effective) buying and selling is what the industry needs to solve. Not squeezing an extra 3 percent of yield or ROI — with 30 percent more effort. That’s an emerging media type approach. We need to see “scale display” as the way we help online display move beyond an emerging media and become a mature media.
3. We need to understand the metrics of traditional media and how they play with the metrics of online. What can we change? What can we give up? How can we make it much simpler to spend much larger amounts of money on our media?

Rich media advertising is a great emerging media type to look at and understand when we talk about scale. Back in the early days, every company had its own proprietary ad formats; some companies were the “expanding ad guys” and others were the “interstitial guys,” and others were the “video ad guys,” and others were the “floating ad guys.” Each company had their own ways of buying and selling the media. Each had their own way of measuring the effectiveness of the media. It was a complete disaster.

It wasn’t until PointRoll figured out how to sell the media at scale, and all the other providers copied its model, that rich media became a mainstream media type within online display. All the providers began offering all the formats and functionality that their competitors offered. The big issue is that they made it easy to buy efficiently, and quickly the percentage of media sold as rich media grew.

Scale display needs to be easy to buy and efficient to manage. We need to make sure that the complexity of an emerging media doesn’t block the success of the entire market. Because I believe that scale display is the way that online display becomes a mainstream media.

(Originally published in iMediaConnection, November 2010) by Eric Picard

In the world of media, content is king. Television networks with the content most people want to watch beat those with smaller audiences. Magazines with the biggest distribution beat those with smaller distribution. And in all of the traditional media, the funding and creation of this high-value content is fundamentally at the basis of their businesses; those who can create it and distribute it to the biggest audience win.

Online media has long been seen as a continuation of traditional media. Yahoo at one point went so far as to hire media legend Terry Semel as its CEO so it could try to replicate the success of traditional media in the online space. This, of course, failed miserably. Similar efforts among the biggest online media companies — those that have focused on building “killer content” in order to attract massive audiences and monetize them — have not done as well as they might have.

Online media is successful when the way the media is generated takes advantage of the power of software. Search engines generate incredibly powerful media because they algorithmically generate content that people are looking for. It doesn’t hurt that they happen to connect advertisers and potential customers at a moment in time that is frequently far down the purchase funnel — a spot that happens to be incredibly rare, has high competition, and therefore drives incredibly high yield. Facebook doesn’t have a human editorial staff writing content; it organizes content written by humans in ways that make that content relevant and interesting to people. And it happens to do so in such a way that it collects incredibly valuable data about the people that both created the content and are consuming it — and can sell the resulting ad inventory for very high yields.

Media companies have missed out on this type of success because they aren’t technology companies. They don’t understand technology, nor do they understand the people who successfully build the most powerful and disruptive technologies. They mainly fail in this area because they believe that programmers and technologists are IT people. They believe that business people, editors, and content creators simply need to tell the techies what to build, and that they will end up with a valuable outcome (which somewhere in there is about creating great content that attracts a large audience that can be sold to advertisers for lots of money.)

Amazing technologists are not “IT guys.” They are brilliant computer scientists who are creative, disruptive, and inventive. They tend to be way smarter than almost anyone you ever meet in an editorial or sales discussion. This isn’t said to diminish the value or intellect of editors or sales people. Rather, these software wizards are simply among the most brilliant people alive. Rather than applying their creativity and intellects to develop smart bombs, encryption software, or a cure for cancer, they have applied advanced mathematics and programming techniques to build better media models.

Sometimes they get it wrong. But every once in a while, they get it very, very right. And when they do, the results are astonishing. They create more revenue as individual companies than some entire industries — or countries, for that matter.

And again — I say all this in no way to diminish the value of human-created content. But simply writing good editorial will only get you so far in the online industry. You might build a great blog or even an associated blog (TechCrunch). And you might be able to attract humans with incredibly well-written or produced content online (New York Times, Hulu, The Onion, Slate). And some of the places where other people’s content is curated well (MSN’s WonderWall) work pretty well.

But the reality of value creation online is where you can apply the power of software to do something unexpected and valuable — in completely new ways. The portals all have the opportunity to do this, but they’ve tended to take an approach to creating media that doesn’t use the power of software to increase its value exponentially. Instead, for the most part, they take a tabloid-like approach, assembling hot topics for their homepages such as, “Details of Tiger’s new estate,” or “How to make a small room look big,” or “What we learned from NFL Week 10.” Yahoo seems to be in rapid decline, and Paul Graham’s essay on why seems pretty telling. I think Microsoft has the right DNA, but not the right focus to pull it off (i.e., online equals search at Microsoft these days). AOL seems to be in rebound, and I’m curious to see what it pulls off in this area — but can it overcome its brand’s association with the early days of dial-up internet?

There are dozens of Googles and Facebooks waiting to be started in the halls of computer science departments across the country. And media will continue to be revolutionized. As will the way that media is monetized.