You Optimise For Users And Ses..., What about your revenue sources? Options

[iamlost]

The intreguing thing about advertising is not that most of us become 'blind' to it's presence but that when we are 'in the market' appropriate ads suddenly become visible.

Thus most ads, particularly those for established brands, are simple reminders looking to catch the attention of someone in current need of that product, service, or information. However, there are exceptions, particularly the introduction of a new brand where consumer 'education' is paramount.

This is most easily seen in TV automobile ads. At the beginning of the model year the ad spots are highlighting the 'improvements', 'gizmos', and selected specifications: educating; later in the cycle the ads have dropped 'facts' instead pushing 'feel': emotion charged reminders.

And just why should a webdev care?

Effective frequency depends on three factors rather than on a single exposure level. These three factors are brand familiarity, message complexity, and message novelty.

Gerard J. Tellis
Effective Frequency: One Exposure or Three Factors?
Journal of Advertising Research, July-August 1997.

So far I've just spoken about 'brand familiarity': teaching the new, reminding about the old. How many of you selling ads directly knew the difference and suggested supportive campaigns? Thought about how cpm and cpc might differ for each? How conversion funnels might best vary for each? Etc.

Next we come to "message complexity". One can not exactly stuff much complexity in the typical text ad. Have you thought about how best to offer educational branding campaigns on your sites? Remember that the ad agencies and corporate ad departments are generally web noobies and most of those who are not only know text ads and equivalent small display ads. They need to be educated about possibilities...do you have any to offer?

And finally, what webdevs love best: "message novelty". Of course you do, simply replace 'novelty' with 'virality'. Exactly.

Once you begin to sell ads directly all the old offline knowledge sets come into play. Have you studied how best to adapt advertising theory and practice to the web? To your sites? Just as many webdevs instinctively optimise both for users and SEs, and some for ppc, they should also be optimising their ad offerings, their major revenue streams. You've built it, they are actually coming...now sell some drinks and fast food and ad space...

[cre8pc]

Possibly related? The failure of ads on a site.

Facebook Ads just in the way is an interesting look at why Facebook isn't pumping out their expected revenue and why Google does.

They're using Facebook to keep in touch with friends. Putting ads in front of them is like hanging out at a party and interrupting conversations to hawk merchandise. "

Avenue A/ Razorfish, a leading online ad agency, bought online ads worth $735 million last year, but only $55 million—a mere 7 percent—went to social-networking sites like Facebook, MySpace and others. Jeff Lanctot, chief strategy officer at Avenue A/ Razorfish, sees Facebook as a communication tool, akin to e-mail or instant messaging. Those are useful things but not great vehicles for running advertisements.

What cracks me up are the zillions of splogs (spam blogs) that rip off content and inundate the pages with ads. It's a failed revenue model but they keep doing it anyway.

[bobbb]

the zillions of splogs (spam blogs) that rip off content and inundate the pages with ads. It's a failed revenue model

Since they are still doing it then it must mean there is profit to be made. What I really wonder about is why people click on those ads... or even stay on the site.

It is pretty obvious that it is junk when you land on a site that has ads on top, left(right), and a search box with only one little paragraph of text (that is sort of meaningless). Even if you do not understand SEO or even know that it exists it is obvious. We throw out stuff that comes in our (snail) mailbox without even a look, we channel surf during during commercial breaks or go for a sandwich or other needs. This is the electrinic equivalent.

I'm sure there are folks roaming these forums that are into this and making money. Hats off to them! Maybe they could explain or shed some light.

[iamlost]

...shed some light...

While I do NOT develop splogs...or even blogs...I do know several people who do, be they splogs, scraped or Markov generated, whatever. The following has been gathered from overheard conversations and disjointed rebuttals of my snarky comments:

What you and I and many/most people see as crap actually invokes one of two responses from those who land:
* the back button
* a click on a link as positive revenue flow.

I am told that a 30% conversion rate is not unknown and 10-12% is quite normal. Given that a web-typical site converts at 2-5% that is an enormous difference.

And Google's bark is far worse than it's bite, possibly because there is little algorithmic difference between these individual crapdev sites and those of some Google service partners. Yes, they do get their domains whacked regularly but typically not before pulling in a whole lot of money.

Autogeneration means they can be spawned at least as fast as they can be whacked. The left hand of G provides AdSense accounts while the right hand takes them away, eventually. And eventually can be quite profitable, so I am told.

[iamlost]

The failure of ads on a site.

SM has a BIG problem, which your quotes reveal: they were almost all created without a revenue model in mind. And then when looking around after the fact, all they can see are the Google and DoubleClick models. Which are not a good fit.

In the OP I was not addressing ads for or by SM sites. Nor even for or by SEs. I was discussing how normal niche sites, particularly information providers (but also product sellers), might look at advertising and adapt it to fit them. That advertising as revenue is at least as important an optimisation concern as search is for traffic.

Webdevs spend an inordinate amount of time SEO tinkering and even more time discussing it. Yet revenue is seen as dropping a script from some ad serving network on the page and hoping. Rather like a fisherman casting a hook out and hoping.

That of course is sportsfishing. A commercial fisherman is optimised to the teeth with fish finders and large nets and contracted markets. It is about time webdevs stopped sports adfishing and investigate going commercial, going direct. Yes it requires a heavy investment: in traffic generation, in analytic analysis and demographic breakout, in targeting niche ad space buyers and selling them on you, your site. It might mean developing appropriate ad campaigns. It certainly will require a lot of client educational development time.

For every hour a webdev spends on SEO two should be on marketing and advertising. That includes conferences, conventions, and networking. It means developing web ad concepts that are not like everyone elses'. It means being a business person as well as a web developer.

[EGOL]

One quote that you seldom hear around SEO circles.... "It is easier to double your income from current traffic than it is to double your traffic."

I have found this to be true in terms of ad format and placement. Those are the two primary variables with low quality sites - they simply focus on getting the ads clicked.

When you move to high quality sites then placement of ads competes with visibility of content. If ads dominate then quality content will attract very few links but if content dominates the ad income drops. The key is to find the balance between link attraction/visitor engagement and being sure that the ads are seen. The emphasis on these two can change at different points in a site's timeline (or success). It is very difficult to understand this. Ad clicks can easily be measured but linkability is not easily measured.

iamlost is talking about the content of the ads and how to engage visitors for profit. Those are multiple variables on top of these.

I don't see these topics addressed well anywhere. It seems that a few people are trying to find their way in the dark but most are simply in the dark on these subjects.

Do you know of any good books, blogs, websites or conferences that address these interesting topics as a primary focus or at least on a frequent basis?

This post has been edited by EGOL: Oct 13 2008, 02:09 PM

[cre8pc]

Since they are still doing it then it must mean there is profit to be made. What I really wonder about is why people click on those ads... or even stay on the site.

Profit maybe? To the tune of millions and possible interest by VC's or acquisition offers? Not likely.

I wonder too about who in their right mind would click on any ad on these sites.

I found the frustration by Facebook and their ads fascinating. They're making a profit but NOT to the tune of what they projected or had hoped for from their ads.

The expectations of SPLOGS are puny compared with what a "real" company is looking for.

[iamlost]

In several of my posts I have mentioned the webdev developing ad campaigns and then 'selling' them to the prospective client ad agency or corporate ad director. And it has been greatly totally ignored. But there really is reason to my madness.

The corporate advertising world is very different from 'our' web world:
* they look different: generally in suits, freshly manicured and barbered, and over 30.
* they judge, very severely, by appearance.
* they live by their Blackberries but take notes by pen on paper.
* they tend to blank at UGC (User Generated Content), to them it is CGM (Consumer Generated Media).
* they constantly complain about the lack of quality web content, well actually they blank at 'content', to them it is 'inventory'.
* they are still uncertain about best integration of video - most testings have been skewed by having been created by either a 'tech' or a 'designer' with corresponding poor performance.
* most are on LinkedIn, some have FB pages but it is as individuals - business/advertising leverage of SM is barely on their agendas.
* and best of all: most are women.

There are a few of very important considerations in the above:
1. they really really really want to advertise alongside the very best niche-demographic-targeted-traffic-drawing content/inventory available. Being best in class is a major selling point.

2. they do not view site visitors as 'users' but as 'consumers'. That is a significant mindset difference.

3. they are adrift in the new web world.

Basically the web is terra incognita and they really can use a guide. So showing how best to integrate text ads, image ads, display ads, video, mini-ad-sites, etc. especially if you can back your suggestions with numbers (testing, always testing) is the greatest selling point of all.

Given they have little practical idea how to effectively harness SM sites directly (SM advertising is typically poor ROI) being able to offer significant SM traffic via your site is a major selling point.

The various advertising conferences draw all levels of agency and corporate personnel: from interns to CEOs. But very rarely a webdev. The greatest missed revenue opportunity of all - cold calling a Fortune 1000 company is, well, a frigid experience; following a conference introduction or lunch conversation with an email, phone call, and draft proposal has a much greater chance of success.

If the ad networks ala AdSense are sufficient - great. If not, you need to get to know the advertising players in your niche(s). And get them happy to know you.

While you introduce yourself to the Advertising World get your traffic numbers up and your traffic analysis done. And did I mention getting your qualified traffic numbers up?

You have got your foot in the door and they keep talking about either what works in print, or some strange Flash creation, or text ad links to some boring corporate 'landing' page... It is time to create some proposals: powerpoint presentations are great, as are storyboards, or mocked webpage samples. Show them the possibilities. Prove the clickpaths and conversion data.

BUT but they are my first direct big client - I have no historical data to show and prove! So before you show up run at least one real life simulation of your proposal - consider it a research cost, an investment in the future.

And provide several gorgeous (looks is nearly everything) bound copies with CD and/or thumbdrive included for later consideration. Great detail is not a good idea: you are providing the idea and selling the service/product. Providing a taste selling the whole.

And you had best have a competent contract attorney available. Preferrably proffer your contract rather than accept theirs. Know your business and be professional, they do and are.

[bwelford]

Just to chip in, iamlost, and say what a great thread this is. In general traditional marketing folk don't have an Internet mindset. Indeed they may often listen to respected and somewhat ancient gurus that were making pronouncements before the Internet turned the world upside down. Things really are very different now and some of the big agencies are finding it difficult to cope when it's difficult to control and push the message.

[glyn]

Hey Cre8,
I'm not surprised that the social marketing models for revenue generation are so lame. The thing that has always got me is that the moment a social media platform starts getting too accurate at the ads it's displaying, the chances are that they are just going to freak out their user base. Remember when Gmail launched and they were accused of sniffing for serving relevant ads based on your emails content.
Yet, if you take a look at the advertising opportunities that are being offered, the profiling is lousy.

What's more, many of these platforms declarations about how they protect your data are so unclear that as an adult with a number of years experience in this sector, if I can't find out exactly what I'm giving up when I install a new application into my account and to whom...then you can pretty much forget others from doing it too.

With regards to the millions of websites that have the same content using a redundant business model, as to the why I think that once everything is in a database it costs probably EURO 20, to do a new website, and that would go some way to explaining why these people continue to flog a dead horse.

Glyn

[EGOL]

iamlost,

This is a great thread. I am glad that you started it and value what you have shared.

I need to figure out how to do these things.

If you have traffic, a quality site, and a contract you still need some method to meter the traffic. If I sell cheese by the pound I need a government-certified scale. What type of metering is acceptable for selling website advertising or traffic?

[iamlost]

If I sell cheese by the pound I need a government-certified scale. What type of metering is acceptable for selling website advertising or traffic?

There is no 'standard' that I am aware of beyond what the two sides agree to in the contract.

A rough outline of mine:
1. Clean traffic.
The very first 'must' is that non-genuine traffic must be accounted for and discounted.
I heart Bill Atchison (IncrediBill): Impact On Your Bandwidth Will Be Minimal My donkey.

2. Visitor number and valuation.
* the type of referer, i.e. direct/bookmark/type-in, backlinks from sites or blogs, SE queries, must valued, i.e. conversion percentage (note: if multiple conversion models each must be valued separately).

* as much demographic breakout as can be inferred, preferrably by referer type.

* clicktrack analysis cross referrenced with referer and conversion.

* all the usual stats: uniques, page views, etc.

3. Page number and valuation.
Each advertiser has various value targets: certain demographic weightings, target term preferences, etc. Thus I break a niche into nine client specific, often campaign specific, page valuations: A1-3, B1-3, C1-3.

The letters are gross relative values, i.e. directly relevant, indirectly relevant, and the rest. The numbers are simply finer filters. And each of the nine levels is given a base cpm value for that client/campaign. And a guaranteed base traffic number.

For instance (example only, not real values) I may sell ad space on 100 A1 pages at 10$/cpm basetraffic 100,000 per month == 1000$. I may guarantee that number such that if traffic is only 80,000 the cpm drops to 8$/cpm or $640.00.

Another example is a combined campaign: I may sell a text ad on those 100 A1 pages at 0.10$ cpc with the link going to a landing page with several other options, i.e. a link to the client's site at 1$ cpc, a downloadable whitepaper at 0.50$ per download. In other words each initial interested click may be leveraged into other revenue options.

But none of it works for long if your traffic is contaminated with bots and scrapers and other dirty traffic. They are by far the main cause of many sites' high traffic numbers (over 80% of some sites' can be crap) which look good when selling to advertisers until the quantity is not quality argument is heard and cpc and cpm rates go down the drain and clients want refunds.

But if the numbers are reasonably clean then it is up to you what ad types on what pages, etc. you want to offer at what rates. Fortunately with compters and databases some quite complex offerings can actually be quite simple to design, offer, adjust, and maintain.

This post has been edited by iamlost: Oct 13 2008, 03:44 PM

[EGOL]

Thank you very much for that detailed lesson on traffic quality, numbers and valuation. I really appreciate it! Awesome information.

[praveen]

iamlost, what a wonderful thread this is and thank you for all the lovely tips!