Web Analytics is considered to be me relatively new field but is it?
One can argue that Web Analytics exists as long as the Web is in existence the only thing that varies is when businesses realized its value and when did we managed to move from simple click stream measurement to more sophisticated data captures that relate to geo-location, time measurement, basic visitors identification (New Vs Returning) etc’. As I am not a veteran I will avoid to give dates and years; nevertheless it is safe to say that is a practice which counts more than 10 years on its back. This is a whole decade and no matter how we would like to play it smart Web Analytics is something that the business world simply knows about and especially after the big boom of Google Analytics where Google provided a free and effective tool for all for its own benefit (ofcourse).
As in every tech product and discipline evolution during the first years software providers provide an expensive technology due to the fact that the inovation required behind the tool is expensive (even not necessarily), the costs to market a new product in the market and build brand awareness is high, but most importantly because during the ‘virgin’ years demand exceeds supply . As we are very well aware today we are in a thriving industry that has more vendors than your mind can remember and more are popping up.
Keeping that in mind, traditionally the way that Web Analytics Vendors have been pricing has been through pageviews per million or in some cases you could gain a software licence and pay on every update in a hosting solution. This has changed with the SaaS model for which you see vendors such as WebTrends who have been traditionally delivering on Premise solutions to push their SaaS service, it is apparent that SaaS works better for clients as well as Vendros. However although in 2009 you could have a WebTrends deal for 25K if you were hosting it, you could take the same deal for 45K on a SaaS solution the argument here is that you pay for the ease of use no need for your technical team to get involved etc’ But do not forget the year 2009; during that year the economies of scale have been very well developed with the hardware, bandwidth and relevant costs to be constantly going down so this 20K markup does not actually justify just the ease of use. (On this point let me clarify that this is just an example and the figures are close to real)
Thankfully during these years our Web Analytics have evolved dramatically from dummy data collection to a practice that delivers insights identifies trends, sets KPIs and proposes actions; plus today’s CMOs know and understand that is all about the informed actions that businesses can take. Thats is where the MVT software providers come into play, who follow again the same pricing models of pageviews or server request or events that an experiment generates. The MVT market although not as old as the traditional Web Analytics it still counts at list 5 years on its back with the technology required being based on traditional Web Analytics and just extending the already known framework further; in short they did not need to reinvent the wheel to make MVT happen.
MVT vendors again follow the same pricing model and deliver a SaaS service. However due to the value of the product and that is still relatively new they can charge a double or higher figure of what would you spent for your Coremetrics, NetInsight or SiteCat solution. Yes there is a value here as your actions will be translated to actual revenue metrics and in a sense you will be able to cover the expenses and be profitable if your team or agency knows what is doing. But still do these high pricing figures make sense? At the end of the day if you have 20% less costs on the tool that means that your investment will be 20% more profitable – huge difference. Google has introduced the last years another free tool called Website Optimiser which provides A/B and MVT testing and frankly few organisations are ready to go in the deeper waters of Segmentation, Behavioral Targeting and Personalization.
MVT is what organisations seem more willing to invest, as they can more easily anticipate the value in it. Nevertheless MVT needs a Web Analytics solution running, given the current pricing costs for an MVT plus an Analytics solution one can easily reach and exceed the 100K figure per year regardless the combination of vendors that looks into (speaking for the industry standard vendors as Omniture, WebTrends, CoreMetrics, Unica, SiteSpect, Maxymiser, Optimost etc’). Again the question remains how is it possible in economies of scale the prices not to scale respectively. Before MVT this price tag could be reached only by Analytics now actually this is replaced by combining two products for this figure.
Then we have how Analytics vendors gain business there are actually only two ways around it, either by direct client sale or through agencies. Now the way that business operate these tools is three fold either through their internal team or via a third party agency/consultancy or via the vendors consultancy services. Every business has its own way of operating but in most cases they will either be on a mature state where they will hire their own analytics team or outsource it; here is where a major conflict comes between agencies/consultants and vendor services as the two parties consultants and software providers have overlaping business interests, the major issue here is not competition but the confusing client. As the MVT consultant will need to sell a service as well as a software provider; where the software provider may already have tried to sell a service to the client – lol!
What Google and Yahoo do is that are concentrated in being solely a software provider and work through partnerships which helps them to easily penetrate the agencies portfolios and gain revenues through advertisement. However the rest of software providers cannot follow up this practice as they cannot gain ad revenues, but the principle remains by building trusted partnerships that can frame the symbiosis between the parties and even more if these partnerships can be tangible translated into a financial benefit to the end client in the form of 30% to 50% price discounts and other means; this would help the vendors to gain substantial market shares and be able to justify why the client should pay for the services and use the the existence of Google and Yahoo services as the vehicle to their service.
Further we really need to see inovation happening, for argument shake the MVT one tag technology it simply remains a myth and still a need in the market. In a large extend most vendors use past technology that does the trick, few things to say for real inovations, the market is thriving and we see that there is more of a straggle how to gain marketshare instead of how can they better position their products and service. I am not aware about the exact figures nevertheless I see a heavy vendor staffing on the services department and I would not be suprised to see them having the same amount of resources allocated for technology as with services.
Why it does not make business sense the current practice?
Just see the relationship in what you can have for a 5 digit figure:
- A top of the line Analytics and MVT solution plus take the vendor consultancy service. – 2X amount
- Outsource the work to a top Agency to run you as many tests as you can possible can with Google/Yahoo Analytics and Google Website Optimiser – X amount
- Build a team of two people 1 Senior and 1 experienced analysts to work for you and use Google/Yahoo Analytics and Website Optimiser – 3X/2 amount
And here we have the argument: the Agency fee is less than the staffing as the agency scales its product (service). However hiring dedicated staff in your businesses has it own advantages and definitely may suit better your needs. But as you can see the full service provided from a vendor simply does not make sense and the funny thing it that only X/2 will go to consultancy where 3X/2 will go for the tool. Therefore having so expensive software solutions simply is questionable if it really make sense for small organisations and if the investment of large organisations if it is profitable enough. In my eyes such an investment should be able to make 10 times its money.
If vendors could provide their tool in a scaled down price through agencies (and some do) then you may turn out with a fourth choice wich would be:
4. Outsource the work to a top Agency use Google/Yahoo Analytics and a top MVT solution – 3X/2 amount.
Here comes the 9/10 rule of thump that Avanish is speaking about from your $10 dollars that you are willing to invest $9 should go to human resources as it is what makes things happen. Even more on the 2010 Omniture summit it was stated that for every $8 spent in acquiring traffic $1 should be spent in converting this traffic; this $1 does not contain the software costs!
Now this is what has lately hit my radar it seems a really cool service Visual Website Optimiser I have not tested yet but it seems very promising and really gives a direction on where thing should be going.
So software vendors need to find new ways of pricing, of making business and scale as MVT is hitting mainstream they really need to find ways to engage more practitioners and clients to their services so that they can gain some awareness on this pool of businesses small and large worldwide that has never heard their name 
Where do we go from here?
We should go OpenSource!
At the moment we have two poles on the one side is the free services from Google/Yahoo and from the other there are the highly expensive paid ones. We really need an OpenSource platform that can work on MVT, Targeting and Personalization there is simply a gap in the market for it and definitely enough demand to make things moving.
Who ever is interesting in working for such a project please contact me.
Evolve or die.
