Is that majority interested in SEO outside of Runet (Ru-Net -> the Russian language Internet) or is English language SEO sufficiently similar?
Is that majority interested in the broader (beyond search) aspects of optimisation? Of marketing?
Webdev is such an intertwining of 'specialty' strands that it is easy to speak either too broadly or too narrowly to an audience's expectation.
Indeed every business/site is sufficiently different in customers, competitors, business model, competence and capability, requirements, corporate culture... there is no one fits all practicable ROI solution... so I guess you need to speak more to best practices, general principles, possible frameworks and processes, etc.
When looking to increase business/site ROI you need to start with some simple questions:
* what specifically do you want to improve?
* what do you need and need to do to get it done?
* how will you know when you get there?
* how can you measure where you are, where you want to finish, and points between?
Note: if it can not be identified, it can not be quantified; if it can not be quantified it can not be measured; if it can not be measured success/failure can not be proven.
* what collateral impact(s) might occur?
* what should/can be done about such secondary effects?
I've enjoyed the discussion so far and have little to add except on the matter of Customer Lifetime Value (CLV), which has been important to me from long before the web. Why have I never remarked on it? I don't really know...perhaps because it focusses on the long term and many/most webdevs have difficulty with short term metrics.
Most are fixated on Visits, Bounces, Click Throughs...
Many calculate gross conversions...
Some track micro-conversions...
Notice the fall off?
And those are all gross averages
; they are traffic statistics, measuring the crowd en masse.
CLF is magnitudes different, in mindset as well as technical requirements.
It is most commonly utilised in ecommerce where customer order attribution is fairly straightforward. However, it can also be used to varying degrees of granularity by non-ecommerce sites.
At the broadest one can consider refering channels as 'customers' and calculate the LV of search, of direct, of social, etc.
Next position in one can consider the LV of sites within each channel, i.e. for search: Bing, Google, Yahoo...
Perhaps the most difficult step outside of ecommerce is to differentiate individual visitors; the methods used range from fairly innocuous (less accurate) to extremely intrusive (quite exact).
Note: with the advent of apps for (primarily) mobile and HTML5 client-side storage the possibilities are an embarrassment of riches.
Over the years I have read things such as direct traffic converts n-times better than search but it is easier to get search traffic...
or Bing converts better than Google by n-times but Google sends so much more traffic that Bing isn't worth pursuing...
When such are common responses to the broadest of conversion segmentation why bother suggesting more granular calculations...
Perhaps the greatest barrier to improved ROI is that most web businesses/sites are so constrained by the cost and time required to simply operate as to be unable or unwilling to undertake the data collection, analysis, and testing required to improve. In other words, they are typical small business. Unfortunately, those with the means, larger enterprises, are as constrained by corporate structure, culture, and inertia.
Edited by iamlost, 13 September 2011 - 03:54 PM.