JayWeintraub.com - Internet Advertising Analysis and Growth Insights

Lead Generation - Quality over Quantity

There was a time in lead generation when all that mattered was the number of leads someone could produce. The price paid per lead was almost exclusively based on the numbers generated, not necessarily how those leads performed for the end buyer. Those buying leads instinctively knew that leads from certain sources did better than others, i.e. leads from email did better than co-registration, but all too often vendors looked out for only their bottom line, charging buyers full price for leads they knew contained, at best, a mixture of sources and, at worst, nothing but garbage.

Today, suppliers of leads to the market still look out for their bottom line and that can still mean acting in a way that isn’t always in the best interest of the lead buyer, but sneaking in the garbage has become harder and harder. The change from quantity to quality has taken several years, but the environment has in fact changed quite dramatically. This is the message that many echoed at the recent invite-only Online Lead Quality Summit being hosted by TARGUSinfo. The validation firm has made a name for itself in the online space, even though this piece represents only a fraction of its overall business. That the company can organize a one and half day event around the topic, though, is proof about the changing landscape.

Perhaps no other vertical summarizes the changes in online lead generation quite as well as mortgage banking. Today, if an online firm says that it generates mortgage leads, a person will be curious what their edge is. But, say that three years ago or more, and that same person will have to withhold from making a snap decision (negatively) about the company. In other words, more than a few supplying leads as the refinance boom started did not earn the best reputation within the Internet advertising industry.

I don’t have the figures, but I would imagine more than a few people decided to enter the mortgage banking field during the refinance boom, either as an independent broker or by joining one of the large institutions. And, these people needed leads. Mortgage banking had already taken advantage of online leads prior to the refinance boom, so scaling operations to accommodate more leads didn’t necessarily require reinventing their business. What it did mean, is that companies made projections based on the number of leads needed, close rates, and the amount per close to determine just how many people they would need in order to meet those projections.

With so many companies going through the same exercise, you had a confluence of events happen, at least in my opinion. It began with a huge flood of demand for leads entering the market; this demand entered the online advertising market but not in a coordinated method. The buyers of leads didn’t know with whom to run; many of today’s firms existed during the lead rush but their reputations had yet to cement. That left buyers in the hands of whoever got to them first; then with the volume needs in hand, these buyers ended up being more brokers than anything else. They would then farm out the deals to whoever said they could produce and the cycle would repeat.

The information asymmetry bred the operational inefficiency that, as the market cooled and competition increased, began to highlight the weakness in the entire system. This ended up with buyers having a lower lead to close ratio than in the past, frustrated employees, upset customers, and a sense that the buyers were paying more than they were receiving. Buyers started to then cut down on the number of companies with whom they dealt, cut back on the number of leads, and implemented systems to better track performance of the incoming leads. Those on the supply side, that both sensed changes afoot and intended to survive the shakeout, began demanding more accountability with those that they worked, implementing technology solutions to improve quality, and in many ways, earning their money.

The system isn’t perfect but enough innovation has occurred and enough external pressures – higher cost of media, greater performance tracking to identify low quality media, and decreased demand – that overall lead generation moved from simply focusing on quantity to quality. Will quality always remain more important than quantity? I think it will, but when another industry develops an insatiable demand, chances are quality will again slip and those in the field look to fill demand. The good news is that while it might take a step backwards, this occurs after taking two steps forward.

Eduventures - Higher Ed meets Internet Lead Gen Part 2

Part One of this piece on the recent Eduventures conference in Boston sets the framework for the information below. It tells of the unique nature of the conference as it brings together, not necessarily by choice, two very different worlds, the world of post-secondary education and Internet advertising. Both the education market and the Internet advertising market have seen high growth and intense competition. Institutions outside of the large for-profits now feel the need to turn to the Internet for more than content delivery. They look to it in order to drive awareness and enrollment growth. Lead generators of course are happy to take their business, but outside of a handful of schools, most remain a ways away from plugging in to the machine which is online education lead generation.

The brief speech I gave tried to help the firms navigate not just the Internet landscape but also the physical landscape, i.e. the small table-top exhibit area overrun by us lead generators. As part of a panel, I had fifteen minutes at most to cover the material, not enough time for attendees to absorb it, but perhaps enough to provide a flavor, so that when they reviewed the physical printouts later, it would make sense. One of those slides is here, and is my take on the types of vendors found in the education lead generation space.

Those at institutions that currently accept leads understand varying pieces of the diagram above. Many already use outsourced software solutions to help them take in leads and track their progress as a percentage of the students move towards enrollment. Two years ago the lead management bubble didn’t touch the others which are very Internet focused. Lead management companies started to see the increasing role that online lead generators play and started to bridge the gap between managing leads and managing the vendors (lead providers) who supply them. That said, they still offer a pure technology solution. At best they act as a silent agent for the schools and, at worst, a gatekeeper that inhibits the marketers ability to help drive enrollments.

As for the other groups, most reading this have a keen familiarity with the other categories, but they remain very Internet advertising specific terms and not common sense to most at the show. Engaging them in the subtleties is the equivalent of getting a non-car person excited over the difference between DOHC versus SOHC. It’s necessary, just not easy. Schools do need to know, though, that certain types of firms will play a role in the stewarding of the brand, while others simply act to get the message out in a pay for performance fashion. For the marketer, the key takeaway is the continental shift that some companies drive to help connect our two worlds.

Among the biggest differences between the firms playing a role in generating leads for schools are a) the services they offer – how active are they in the brand process, what role they play with respect to banner, landing page creation, etc. and b) how they drive their traffic. This next graphic applies not just for education lead generation but for other areas of online advertising. It focuses on the latter point, the traffic. The chart presents nothing new to those who have gone through the Internet advertising learning curve, but for those who haven’t, the relative positioning of the channels and the terminology take incorporating into their vocabulary.

One channel not listed but one that proved viable for many schools are online education directories, not necessarily the large ones like ClassesUSA or eLearners but some more niche sites that focus on specific verticals. As schools begin to increase their ability to handle incoming leads, they can then move to other areas of online marketing. The most sophisticated schools have the ability to take leads from each channel, qualify them, and pay for them appropriately depending on the enrollment rate. Not every school at the conference will use or want to use lead generation for the marketing of their school, but many institutions, including traditional brands, have turned many of their degree programs into profit centers and want to keep that growth.

More and more schools seem to feel the need to actively market, and it’s that need, plus the opportunity to hear from institutions large and small that have undergone similar challenges that attract them to Eduventures. It isn’t, however, an Internet advertising show. PPC is often the last thing on a 100+ year old institutions mind as it is trying to convince its Board of Regents that offering a degree online makes sense. The conference does provide a closer look to the behind the scenes challenges current and future Internet advertising clients face, and we as marketers benefit greatly when we listen to their process for becoming a client. As was the case last year, it was a worthwhile show if you were looking to understand a market and not just close a deal.

Eduventures - Higher Ed meets Internet Lead Gen

Last night, I returned from Boston. This week marked the third year that Eduventures, a market research and consulting firm for education institutions, has put on their Competing in Higher Education tradeshow. With only a handful of lead generation firms at the show last year, yet fifty or so for-profits, it seemed like a rare find. Naturally, that didn’t last long. Instead of their being three or so advertising companies, this show almost ten crowded the small table top display area not to mention the others in attendance. Chances are the hundred or so representatives of smaller, regional universities or their counterparts from a few prestigious brands did not know what to make of the Internet marketers that descended upon them.

Eduventures’ Competing in Higher Education marks one of the only shows where the world of running a university meets with online advertising. Those from institutions attend because they run programs that target the adult-learner, one working full-time who has either completed some of their education, wants more education, or education for a new career. Attracting and servicing these students requires different operations than those required for the traditional, full-time 18-24 year old students. As they look to grow the adult-learner student base, it makes sense that these administrators have a dialogue with the online marketers, as the skills required to market profitably online require such specialization that for the schools to try and do so internally will become a distraction to their core business of educating students.

The accountability of advertising online has thus far made a good fit for growth oriented institutions such as AIU Online, who manages the key aspects of their brand in-house but relies on the performance marketers to show their message and deliver leads. Unlike more traditional schools, AIU effectively bridges both the academic world and the corporate world. Most institutions do not have their foot in the corporate world. They have not structured their processes the way a company might, and it will take time before these institutions can embrace Internet advertising in a manner that meshes with the tactical execution services offered by the lead generators. Most others, though, are still far away from being able to accept a lead.

The lack of immediate clients did little to deter the lead generators from attending even though virtually all of the presentations overlapped little with their area of interest. The majority of speakers were from institutions, and they covered topics ranging from how they market their programs in general to how they launched online degree programs, having not done so before, to how to change management techniques helped turn around decreasing enrollment. For the non-educator, like me, their stories and the presentations covering the market trends were at least more accessible than the equivalent presentations I attended in October of last year at the Mortgage Bankers Association trade show. I imagine though, that some in the audience felt as I did ther, when it came turn for one of the few non-educators to speak, me.

Some of the other lead generation firms would have preferred to see a panel or roundtable, anything but just one Internet marketer speaking. Given the economic impact the lead generator’s have, it’s one I suspect will happen at next year’s show. As it stood, I had the difficult but enviable task of covering online advertising to a show where a small subset of schools not only embraced lead generation but probably knew more than I, while the other half did not fully grasp the difference between paid search and organic. That other half included not just educators but also investment bankers that recently purchased small regional schools looking to run them like businesses. Neither the educators nor the bankers would find much value in a talk that tried to explain keyword insertion on the search engines or trademark bidding policies.

To make the most of their time, a talk to this group should have some practical, tactical value. If the information stayed too high level, it does not help them feel better armed when they begin planning or reviewing their online strategy. Often this means providing either a case study or a step by step guide that those interested in online marketing could follow. Both would have added value; each was like giving directions. In order for directions to have value though, a person needs a map. Directions without a map or directions on the wrong map both lead to the wrong destination. It was “just enough information to be dangerous” as one person put it. For any interested in some of what the presentation covered, I have a follow-up post coming. Long time readers will recognize certain elements as this one too contains an education lead generation specific “Venndor” diagram.

(And because I have some domainer traits at heart, I had to register "venndor.com" and "venndordiagram.com".)

Strategies of online education lead generation

The last piece I wrote about online lead generation covered the basics of lead generation and arose out of an agency/broker’s desire to get a CPA price quote for an unnamed school. Instead of simply saying “$50”, I proceeded to write out a lengthy reply on why quoting a single number poses a challenge. Today, we look at a related topic - the various strategies lead generators employ to drive leads to a school. What makes this topic so interesting is that as recently as five months ago, a few on this list were considered on the fringe. The increased cost of traffic to lead generators and desire for schools to focus on a narrower band of providers has opened the door, though for some, once fringe, concepts to make their way into the mainstream.

The “fringe” activity in this case is the desire and push by lead generators to promote more than one school as part of the sign-up process. Three companies have made the greatest in roads – Quinstreet, ClassesUSA, and Nextag. In the case of Nextag, they only promote education in this fashion, whereas both Quinstreet and ClassesUSA offer schools several different ways to generate leads. Below are the four most common mainstream strategies used by lead generators. In this piece we focus on the presentation of the school as opposed to whether visitors came from search, display, email, etc.

1) Single school landing page

• Exclusive lead, i.e. a student sees one brand on the landing page and can sign up for only one school.
• Form fields usually present; strong call to action; an information gathering page rather than an information giving page.
• Generally the highest quality but also among the most expensive.
• Viewers tend to see either a general landing page for the school,l or among the more advanced lead generators see a degree specific page, e.g. Criminal Justice based on prior intent, i.e. a click on a tab on banner or search query
• Some lead providers will, upon completion of this type of page, show a follow-up page with other schools; some schools will bar this, seeing it as leveraging their brand for a competitor.

2) Directory style page / Homepage placement

• Among the oldest style for generating leads.
• No form fields present; limited if any call to action; a focus on information giving.
• Many schools co-exist on a page that typically look to provide users a way to navigate to a particular degree / school combination.
• The non-sales, exploratory approach means that viewers will often convert on more than one school.
• Seen as a clean and often preferred way for vendors to generate leads.
• Easily digestible and as a result historically the type of page show to schools by vendors when attempting to work together even though it drives a smaller percentage of leads.

3) Multi-school “Shared Lead” page

• Credit goes to ClassesUSA for the nickname
• Marketing message driven page; focus on information gathering.
• Similar to the exclusive lead form except more than one school is presented on the page
• Unlike a directory page, focus is on a user completing a form rather than navigating deeper into the site.
• Nicknamed shared because after a user completes the form for a particular school, the page presents the user with other schools that offer the same or similar program, storing the user information so that they simply must click one or two times to be submitted to the next school.
• Students tend to sign-up for more than one school
• Some schools have seen only minimal degradation in enrollment rates by a switch to the shared model; whereas others have seen sharp decreases.
• Gaining increased acceptance among schools, albeit reluctantly.

4) Forced shared lead page

• Currently only used by Nextag and allowed with them given their size and clout in the mortgage lead generation business.
• Page mirrors their mortgage form in that users complete information in the first part of the form and then are “matched” to schools on the second part
• Students will sign-up for more than one school (assuming a degree of interest is selected that more than one school offers.)
• Lead degradation occurs, but it’s not linear, i.e. a lead shared 3 times does not perform 1/3 as well. Closer to one half.
• Most controversial page as it presents the least information per school and does not offer a way for a user to learn about a school and select just that school.
• Additional controversy and push-back from some major institutions due to its commoditizing education. With the schools being given little to no room for differentiation, such forms have the potential to impact negatively the sector, especially if others jump on board. Users will lose out and the page will be seen by accrediting bodies similar to co-reg being run on incentive sites.
• The first type of site to have volume caps reduced and be seen as expendable.

In Effectiveness in Lead Generation and Finding the Sweet Spot we hit upon a few of the reasons why mortgage has scaled so easily and the challenges that online education faced in doing the same. What we see here are companies breaking free of those constraints, but the results of these new formats will determine its long-term success, of course. The best bet for vendors is to treat these strategies like a portfolio and have diversity. The new formats might have strong growth but they are also the most volatile.

Lead Gen Support Business Idea

Like many creative thinking types, I often have more ideas than I know what to do with; some I would like to think I create someday; others just take up space and cause mental irritation because I know I probably won't get to them. Here is one of those -

Targus Info has made quite a name for itself and quite a business providing data validation, their services becoming the Akami of data validation for lead generation companies. I'll have to ask Dave over there if their service offers standard USPS look-up, but that's another topic.

What would help, and Targus could do this, but most likely a company such as Datran looking to expand their revenue stream should offer this, is an email validation service. This is nothing new; when I did email collection through a registration process, I wanted a better system for validating names. It's pretty easy to see if the domain exists and has a mail server set up on it. Finding a way to determine whether the actual address works and should count as valid is something else.

The idea that would seem to work is for Datran or a similar firm to allow others to send over an email address and have them score it for them. The first level of scoring would be whether that address exists in their database or ever has. They would them help shed light on that name by providing any deliverability information - name is valid and receiving email; name was in database but unsubsribed (meaning it is still valid); name is in database but not active (not opened email); name is a spamtrap do not send, etc. With their experience, they could probably add value to names that aren't in their database as well by applying learnings across names as a whole, e.g. whether certain strings in the name suggest it to be fake, or the structure of it to be good (the lead gen provider might pass in first and last name too).

Datran or others probably wouldn't want to share too much of the specifics but they could help come up with a score that would let those who just received the email gain some insight into the validity of the lead. Ideally, some third party would do this and Datran along with the other big mailers would participate, making their data accessible to the third party. The subscription revenue would be split among the data providers (with protections so that their data wouldn't be abused, competitor safegaurds, etc.). Given the forming a single organization (like Targus) who has access to the data owners (ex, Wireless carriers) takes more time, which is why the data carrier (Datran) could do this if they wanted.

Basics of Lead Generation

The inspiration for this piece came from the following email, which said, “I am setting up an online media campaign on a cost-per-lead basis for an online University. I am working on a final budget and would like to come up with an average CPA for this campaign… Please contact me to workout specifics.” Unfortunately, the specifics did not include a client name, but it will require an RFP. Not coming from the agency world, the thought of an RFP to run an online education campaign seems incongruous. It might have made sense in 2000, when much online education lead generation took place via banner ads and not vendor hosted landing pages, in today’s market, this approach just seems anachronistic. I just can’t imagine a school trying to enter this market without the industry experience. Part of me is waiting to hear that it’s actually someone trying to promote University of Phoenix.

The point, however, was not to criticize the approach but to help answer the question of the average CPA and volume factors. In general (with respect to online education):

• The average CPA depends on many factors. One of them deals with the school itself, and that is whether the offer is online only, ground only, or both. If it’s both, one factor is how many campuses they have. Online programs tend to pay slightly lower than campus ones, and programs with greater geographic restrictions will also cost more.
• Course selection plays a role not in necessarily in price but in volume. Schools that offer a larger variety of courses will convert users better than ones with a narrow selection. Ones with only a handful of courses, and especially if niche, will be asked to pay more.
• Form fields play a role too. Forms that require only the bare minimum convert better and can justify a market competitive price. Those that ask more personal and less common questions will have a lower conversion rate, lower lead volume and cost more per lead.
• Who hosts the form is yet another factor. Some schools do not let the vendors design and host landing pages. This in my opinion is a mistake. In the beginning, firms could do this and still get volume. Specialization is the name of the game now. Companies specialize in obtaining traffic. The specialize in landing page design, in optimization of pages, in optimization of which pages go with which traffic sources, and so on. One page fits all will have a hard time competing in a one-landing page will not fit all world.
• Taking the above into consideration, most vendors can quote an average CPA, but the smart ones now focus less on CPA and more on enrollment cost. Many vendors – those with affiliates for instance, more generally those who have various sources of traffic, see different performance across each. One price fits all means vendors must overpay for some and underpay for another. Again, this is because not all traffic performs the same. It’s the vary reason Google has their Smart Pricing, charging advertisers different CPC’s for the same keyword based on the source of the traffic.
• A factor that this planner should take into consideration is whether the offer is a school that is already being marketed; if so, it will be difficult to get placement unless this person’s company is taking over all placements, ala Advertising.com and Apollo. Generally, only a new school that hasn’t been online will get consideration with the stated restrictions – one price, potentially one landing page, RFP.
• A point to this planner is to have them prepped for the fact that most vendors will ask to host the form and transfer data to the school; it’s driven not just because of specialization but that many vendors will want to integrate the school into their existing education portals
• As for potential volume, the largest of the online schools process more than 200,000 leads per month. Schools that have low budgets and lead caps will have a hard time getting an audience. This school will want to be able to accept at least 500 leads. The fewer leads they accept the more it will cost per lead.
• So, what does all of this mean if pressed for an average CPA? Typical schools pay out - $12 to $20 for a shared lead (similar to mortgage model); $30 to $40 for average exclusive leads, and $45 to $55 for higher quality leads (better student to enrollment rate). But, I strongly suggest and encourage being able to track not just leads but enrollments and as granularly as possible.

Most fascinating to me about the email request for information is that it hints at a process that runs counter to today’s right pricing trend. We’ve started to hit a point where even the “shadier” vendors understand that quality impacts price, so it’s easy to take for granted that others might not understand the dynamics of a given vertical. I have assumed the above was equivalent to CPM, known by all. In case it isn’t, may the above help.

Components of Conversions

Being focused more on the lead generation side of online advertising I often view the world myopically. That was the case last week until I received a question that warranted some additional perspective. Someone asked, “How important are conversions that were initiated by a search, but actually converted maybe 2-3 days later? Do you measure this and include it in your search cost per lead calculations?” These questions make up an important topic, one that many in lead generation and other areas of direct marketing take for granted. I know I have, which is why the components of conversions are the focus of this post

Components of conversions are really the study of behavior. It is part of a broader dialogue that deals with understanding how customers make their decisions and using that information to communicate more effectively. The challenge, and we try to solve, here is providing a framework that can be applied to other industries and verticals as they look to make sense of this behavior.

Granularity and immediacy of feedback has long been an Internet advertising selling point. Companies do not have to wait three to five weeks to determine if one segment did better than another or whether that particular buy made sense. Even the Internet, though, offers different levels of feedback. We can plot these levels on a spectrum from immediate feedback to delayed feedback. Those businesses that exist on the immediate, are those that know exactly what they make within a short period of time, say a few hours. On the other end of the spectrum, would be companies where acquisition and an understanding of a particular click / customer’s worth occurs after a lengthy delay, e.g. an Amazon.com customer or new car purchase.

Lead generation, as an example, falls toward the left hand portion of the spectrum, very close to immediate. This implies that calculations for search or other acquisition campaigns do not need to take into account delayed transactions. Determining whether or not delayed transactions play a significant role at all is something that, as mentioned before, those in the online ad space often take for granted, or don’t consider, given the singular nature of their product. For everyone else, presented below is a way to help assess where their product / service fits on the spectrum if they don’t know already (or just want to double check).

Example 1: Components of a Conversion – Solving the latency issue by determining whether to account for the existence of delayed conversions in a conversion.

The four major aspects to any given transaction:

• Time commitment / Comparison behavior– length of time needed to complete the transaction. Is it spontaneous or deliberate? For example, on one end would be a lead form where it takes two minutes or less to complete, and no real consideration whether to do it; on the other end would be a big ticket item where people will often think it over, do comparisons on it, etc.
  

   
• Financial commitment – how much money is required. Is it just your attention and personal data, or do you need at least a credit card, perhaps a loan where there are recurring payments with an often lengthy pay-back period.    
   
• Need for Permission – level of input needed by others in order to complete the transaction. Is it something you can do yourself or do you need to consult with a parent, spouse, etc?    
   

• Repeat Usage – the likelihood that users who convert, will convert again. Are you a single serving site or someone that maintains an ongoing relationship? Incentive promotion sites see people once; Amazon.com customers order from there again and again, on average.

The components above help illustrate why most lead generation sites do not need to account for delayed conversions and why performance data from search engines can be used as-is. More specifically, with most lead generation campaigns: a) Time Commitment is low (it takes two minutes); b) Financial Commitment is low (no credit card, only common information); c) Need for Permission is low (most fill it out if they decide they want to); and d) Repeat Usage is low (people who fill out the forms tend not to bookmark or come back later).

While lead generation is not easy, it is at least straightforward. Unfortunately, as the assessment above illustrates, that does not hold true for sales based businesses, e.g. electronics and other larger items. In these cases, web analytics play a critical role as profitability off initial click becomes near impossible to determine. Companies must track and really learn their customers. It will mean monitoring online and offline actions, but in the end it will lead to smarter spending. The good news is that the further right you get on the spectrum, often the more you can make per person, so those in lead generation looking to increase their yield can start moving further “right.”

Google versus Yahoo by the numbers - Part 2

This post, we continue the fascinating look at performance differences between the two engines in “Google versus Yahoo by the numbers – Part 2.” In a more ideal world, other search marketers would share similar data so that we could all help keep the engines honest not to mention our sanity in check. In the interim, enjoy this other perspective of comparing Google versus Yahoo.

Similar to Part 1, the data in these examples comes from lead generation campaigns. Again we compare Google versus Yahoo, but this time, instead of line graphs that look at the engines’ conversion rates over a period of time, we take a more holistic view for quality. The charts below look first at Yahoo Search Traffic, then Yahoo Content Network Traffic, onto Google Search Traffic, and finally Google Content Network Traffic. Each chart depicts traffic in three buckets based on the conversion rate. Traffic above a certain conversion rate is considered great (green); traffic below a certain conversion rate is bad (red), with traffic in between the two as fair (yellow).

Chart #1 – Description: This chart shows Yahoo Search traffic performance during a two-week period in February. It is a period where, for these particular campaigns, Yahoo performed below their metrics for most of 2005.

Chart #1 – Explanation: Yahoo Search traffic consisted primarily of “Great”, which made up 44% of all visitors with Fair being next at 31% and Poor being 25%. Overall, the traffic is fairly balanced, and were we to do a comparison between this time period and November of last year, the biggest difference would be the percentage of Poor traffic. If you remember, it was the overall decline in Yahoo conversion rates that prompted this series of analysis in the first place.

Chart #2 – Description: This chart shows Yahoo Content Network traffic performance during a two-week period in February. Last week’s comparison did not include data from the Content Network.

Chart #2 – Explanation: This one is a little harder to read but the color dispersion tells the whole story. There is a strong amount of Great traffic (43%), but much more Poor traffic than was seen on the Search side. More than half of the traffic qualified as Poor (56%), and surprisingly virtually none fell in the generous Fair bucket. The takeaway here, is that there are some high quality content sites, but certainly not all. It shows why paying the same click price per partner site is not a good deal and the benefit of right pricing on a granular level.

Chart #3 – Description: This chart shows Google Search traffic performance during a two-week period in February. It is a period where, for these particular campaigns, the engine actually performed slightly better than their metrics for most of 2005.

Chart #3 – Explanation: A sea of green, that is what Google Search looks like for these campaigns. More than two-thirds the traffic qualified as Great (68%) as opposed to Yahoo’s 44%. The levels of Fair traffic were close – 24% for Google and 31% for Yahoo, which means that it’s the Poor traffic where Yahoo had more – 25% to Google’s 8%.

Chart #4 Description: This chart shows Google’s Content Network traffic performance during a two-week period in February. Last week’s comparison did not include data from the Content Network.

Chart #4 – Explanation: Another interesting chart, especially when compared to Yahoo Chart. Google content had a much lower Great percent than Yahoo – 11% to Yahoo’s 43%. Instead of Great, the vast majority of the traffic, 64% was Fair, which might not look so good, but consider that Great plus Fair equals 75% with Google Content Network and only 44% with Yahoo Content Network. The big difference again was the Poor component – 25% with Google but 56% with Yahoo.

This makes two posts in a row where performance data points in Google’s favor. Objectively it helps validate the engines lead in the search race and shows the value of moving towards a right pricing model. But, just because Google converts well and charges variable rates in a more sophisticated manner than Yahoo doesn’t mean they work better for marketers. This might sound confusing, but it’s because of Google’s efficiency that making money on them becomes tougher. There are simply fewer gaps of opportunity, fewer areas where pricing is below the maximum market value. As a result, you get good traffic, but it costs you, often to the detriment of profitability, at least as far as immediate marketing goes. Hats off to Google there for extracting so much out of so many, but I’d personally settle for a little bit more inefficiency, or at the very least their stock price to rebound.

Apollo Advertising.com Strategic Partnership

Apollo and Advertising.com enter strategic partnership.

While not quite new news, the partnership represents potentially one of the most  significant changes to take place in the lead generation space. I did not write about the deal when it first happened because my primary duty is to my employer and its clients, one of whom is University of Phoenix. Being the person to break a story does not benefit me, especially if it could mean straining a valuable relationship. Having already made one mistake involving writing about a client, I wasn’t going to do it again, and certainly not with the 800 lb. gorilla of for-profit education. Unfortunately, the past two weeks has not shed any more information on the deal, but the announcement has reached all stakeholders, which makes now an appropriate time to discuss the deal and its implications.

I first heard about the Apollo-Advertising.com arrangement February 28, 2006 when a sales person spoke to a school that had just been given notice by Advertising.com. Later in the day, someone found an article in MSN Moneythat, like most that day, talked not of the deal but Apollo’s conference call with analysts where the company announced it will miss second-quarter Wall Street estimates earnings. Apollo Group has a history of impressive growth, but its stock has taken a beating recently, having gone from the high 70’s to the low 50’s in the past six months.

In addition to announcing the revenue shortfall, the call outlined the strategies being implemented to continue their dominant growth. One of those strategies involves “an agreement with Advertising.com which operates the largest interactive advertising network in the world.” University of Phoenix has “entered into a highly strategic relationship with this wholly-owned subsidiary of America Online in order to gain an even greater advantage in the highly competitive education space.” And, that, “The combination of Advertising.com’s world-class optimization technology and front-end combined with our capital resources, qualifying center and sales force will give us the potential to continue to lead the industry.”

Partnerships such as these don’t get mentioned in analyst calls by billion dollar companies unless it means something, and here are the two key takeaways:

   
1. Advertising.com will in many respects become not just the agency of record for University of Phoenix but an extension of their marketing department. All leads will go through Advertising.com – even companies that considered themselves competitors, such as Quinstreet and Nextag, must now work through Advertising.com. They will be the point of entry for leads and provide the reporting for other vendors.
   
2. Advertising.com will work only with University of Phoenix and no other education clients for the duration of this deal.    

With respect to the first point, all leads flowing through Advertising.com, the ad network now enters the lead management space currently occupied by companies such as Datamark, Compass Knowledge, and CUNet. Lead management is a tricky space for someone with traffic to enter. Companies that play both roles – technology and traffic, can, if they are not careful, send the wrong message about their intentions. Will they allocate leads fairly (given that they have their own quotas they want to hit)? Will they try to poach affiliates? Those are just two of the common questions in such scenarios. The good news is this relationship has the potential to do the right thing as well as let the good guys make more. Nothing is ever perfect, so there will be some hiccups along the way, but the team at University of Phoenix wouldn’t have entered into the deal if it were going to undo years of trust in the vendor marketplace. Some affiliates might leave  heir current network to go to Advertising.com - they will want to work closer to the source and get paid what they feel they are worth - something that other affiliate companies have not always embraced, but I would expect University of Phoenix to have discouraged the company from any direct recruitment.

Interestingly, Advertising.com was not the only company that was courted for a closer relationship. Phoenix had hinted to several other companies that it was looking for a company willing to dedicate itself more fully to them. It seems that many companies would have fought for the position that Advertising will soon occupy; yet, that wasn’t the case. And I think understanding why is best done by looking at the types of companies that have been able to drive the largest number of leads for Phoenix – lead generation companies and ad networks. Lead generation companies are ones like World Class Strategy and LowerMyBills. These are firms that own the entire pipeline – they design creatives, host landing pages, and manage media buys for those creatives and landing pages. They are experts in their vertical, and in the case of education, doing it right means the company tends to only serve that market. Were one of these companies to accept the deal that Advertising.com did, it would dramatically alter who they are as a company and most likely end in failure.

Ad networks on the other hand focus on just one piece of the pipeline. Their core competencies are not designing and testing creatives and landing pages. They sit in between advertisers and publishers, focusing on the very specific problem of matching ads to placements. Instead of finding placements that work for a particular type of ad (which is what lead generation companies do); they act agnostically – a client is a client is a client, and an ad spot is an spot is an ad spot. For an ad network to do an exclusive for a category does not dramatically change their business; they just have a different ad mix from which to optimize.

I can’t say I would have guessed either to happen – that Phoenix would chose an exclusive gatekeeper or that Advertising.com would be that company, but now that it’s done, it makes a lot of sense for both. From the media side, success in lead generation requires the ability to reach a lot consumers, which Advertising.com has, but, more and more, it requires specialization – creating the best ads, landing pages, and combination of the two. That is something Advertising.com does to some degree, but it's not a core competency and lever to success. With the continuing influx of brand dollars, it doesn’t necessarily make sense for them to shift their business model and enter the vertical lead generation space. Also, their volume for University of Phoenix has stayed flat or gone down over the past two years, not increased. And that has probably been the case for their other education clients as well. They still make up a significant number of leads, and high quality leads at that, but maintaining a position of leadership would only get harder and harder. Advertising.com is, at its core, an analytics and process-driven media company. They understand how to handle a large number of relationships and extract the most from these variables. It was this that allowed them to carve out a position of leadership in the messy and complex ad space. And, it makes them well suited for this task.

University of Phoenix is not a media company. They are the most efficient and advanced post-secondary education company. That efficiency allowed them to carve out a significant piece of the media world by creating an offer that outperformed so many others. As the world of Internet advertising has become more competitive, and more fragmented, it has became harder for them to stay as profitable. University of Phoenix could in theory hire more Internet marketing talent – both in operations and technology. Or, they could just outsource those aspects and continue to focus on their expertise – running the nation’s largest university. And, that’s just what they did. Granted, Advertising.com takes a big risk with this deal, they essentially close the door on several long-time clients, but now they get some upside on all of Phoenix’s leads rather than on just a small percentage of those leads. And, chances are that the other leads they just inherited more than makes up for the combined leads of the other schools.

Most importantly though, the University of Phoenix-Advertising.com deal sends education head first into right pricing, something that had been sorely missing and a topic I’ve written about previously. It also focuses on strengths including those of the other lead generation companies. They will still get to do banner and landing page design, branded media buys and search arbitrage. Affiliate marketing companies can still recruit and arbitrage. So, while this could turn out to be a disruptive force in lead generation, i.e. do others follow suit, it also signals incredible opportunity – not just for Phoenix and Advertising.com but all players, including the other institutions.

Overall, I hope other institutions let this play out before making any major decisions, such as forming a similar alliance in a purely reactionary move rather than market driven one.

Economics of Lead Generation

Economics of Lead Generation - The Price Fallacy

This post stems from work I did for the March 9, 2006 DM Confidential. You may view the two part article here and here. Many thanks to those who have already read and commented on these pieces. Presented below is a condensed version. If any would like a hard copy, white paper version please let me know and I will circulate.

History of commerce tells us that the more we buy the cheaper it should be per unit. Shopping at places like Wal-Mart, Sam's Club, or Costco illustrate this and have trained consumers to expect to pay less per-unit the more units we buy. For example, a single soda might cost $.60 a can, a six-pack $.40 per can, and a dozen $.30 per can. This per unit cost reduction happens through economies of scale. The stores can sell it to us for less because the manufacturers charge them less. The manufacturers charge the stores less because their cost per unit drops the more they make, and big orders mean larger amounts of money coming in more digestible and predictable chunks.

Lead generation as a market does not, however, follow this pattern. On the whole, the more leads you want to buy, the higher the cost per lead will be. There are no economies of scale for an advertiser at higher volumes. What's more, the price per lead over time does not drop; there is no Moore's Law with lead marketplaces. The economics of lead generation mirror more closely those of scarce resources, like oil extraction.

There is only so much oil in existence. The question around oil is not how much is left but how much can be produced, i.e. the supply. In a given reserve, the early oil is the easiest and cheapest. Costs increase as the reserves empty out. This is the key to understanding lead generation pricing. The easiest and cheapest leads come first. The more you buy (extract) the more it costs, as leads are in many ways a non-renewable resource. Unlike oil, we will not use up every lead – the constantly evolving economy and population insures fresh supply. The demand for new leads, however, laps the natural market growth, insuring costs per leads will both increase over time and as volume for leads increases.

With the respect to the economics of lead generation, the important thing to remember is that buying leads is not the same as buying fixed price media. Were you to buy media in bulk on a fixed price, you would get a price break - that is guaranteed revenue for the owner of the inventory, and they would lower the price in order to secure the revenue. Leads, though, are not guaranteed for the seller. Each lead must be earned, and the incremental cost to acquire a lead increases with each one for the sellers and thus the buyers. Over the past six years the market has shown this to be the case across countless verticals. When companies ask for more money for increased volume they do so as a reflection of their costs not in attempt to fleece the lead buyers.

To understand lead pricing, you must  understand 1) how lead markets change over time and 2) the cost curve that a particular lead buyer should expect. Presented below are those two graphs. The first is a view of a lead market over time containing the costs for both buyers and sellers. The second is from the perspective of a lead buyer (the advertiser) and looks at cost per unit as it relates to volume of leads.

Explanation of Graph 1

Note: The red line represents the average cost per lead paid by a buyer in a particular vertical. The blue line represents the average cost of acquisition, i.e. the cost of media; the difference between the two is the profit.

• Stage 1, “Early” – when just introducing a new type of lead into the market the cost per lead an advertiser needs to pay is at its lowest. Latent demand exists; the audience has not been exposed to the offer, and it does not directly compete for media space, nor are many sellers running it. The cost per lead for the lead seller (the publisher) is low but not at its lowest point as there is some economies of scale they will experience as they learn how to market the offer effectively. This period is the “fresh oil” for buyers and sellers.
• Stage 2, “Mature” – the offer has proven itself in the marketplace; this is the point where great strides in volume are made as more and more sources of traffic are used; prices begin to increase as the offer becomes prominent and it begins to compete against other offers for media space. Lead seller costs can decrease during this period as they understand its performance, e.g. where to buy, keywords, landing pages, etc. Towards the end of this period though, both costs for the buyer and seller begin to rise. This period is where “peak production” is reached and a flood of additional sellers will enter the market.
• Stage 3, “Saturated” – costs for the lead buyers begin to rise steeply as the market saturates; the demand for leads outstrips the production rate of supply; incremental costs for leads increase dramatically as buyers and sellers compete heavily against other offers in the market place, and each other; performance declines due to overexposure, which only increases the difficulty the offer has competing for, not just media but, seller attention. Margins for lead sellers shrink, their costs rising proportionately faster than those of the buyers; this is the period of final extraction from the well where vast quantities might exist but in a manner that costs more and more to extract. The lead buyer must focus internally and look for areas of fragmentation and differentiation to stimulate supply and/or reach the remaining supply economically. The seller must try to become more efficient in order to compete for inventory and among other sellers for the same offer. Think Hungry Hungry Hippo; it’s not pretty.

Explanation of Graph 2 – this graph looks at the cost per lead as a function of volume of leads desired by a specific buyer. This graph illustrates why a greater number of leads will cost more money per unit. As mentioned in Part One, this graph is not speculation or fleecing by sellers (publishers); this is the established trend that has emerged across all mature, high volume lead verticals.

• Phase 1, “Low Hanging Fruit” – buyers of leads in low volume tend to pay more per lead than buyers of similar leads that buy in bigger amounts; as buyers increase their capacity they can experience some economies of scale; lead sellers charge more for small clients because of their internal costs and the opportunity cost, i.e. the seller needs to earn a higher margin per lead to cover their fixed costs and to make up for not working with a larger client; as buyers begin to scale, though, they will achieve both internal and external economies of scale resulting in the ability to pay a lower cost per lead on all leads bought; as they grow, they enter into the “Sweet Spot.”
• Phase 2, “Sweet Spot” – during this phase lead buyers reach the optimum balance of volume and cost per unit; companies can scale the number of leads without significantly increasing their cost per lead; here the number of leads purchased matches available demand; it is comparable to the period of oil extraction where production reaches a constant flow and stable costs; as the desire for greater lead flow during a given period increases, companies enter Phase 3.
• Phase 3, “Chasing the Tail” – a section characterized primarily by diseconomies of scale and a non-linear relationship between cost and volume; during this period, as volume increases, it passes the optimal point, meaning there is no added benefit for both the buyers and sellers; the easy leads are gone; there is competition with other advertisers for the better converting inventory; the offer is saturated, and sellers have other, potentially more lucrative uses for their resources; with eroding performance and margins, prices must rise to finance the high production costs; peak production will be reached and buyers will find themselves hitting a volume wall where increases in cost yield negligible increases in volume.

The market view and buyer view, Graphs 1 and 2, are designed as two complimentary ways to explain, visually, how costs for leads rise over time. Both time and volume play similar roles. Lead prices will increase over time as the easy leads get taken early and auction effects come into play. As volume needs increase, more and more of what’s left are the harder to reach leads, combined with more expensive media, and increased competition among sellers to reach the shrinking pool of people. All of which add further price pressure as sellers have an incentive to work on other industries. It’s a cycle that cannot be escaped, having already played out in countless companies in a variety of lead verticals. Increasing prices for leads are as the Matrix’s Smith says “inevitable.”