Quality

PRICING VISITORS

If the end goal of a web site is to generate a profit rather than merely an audience then it's content should be dedicated towards those visitors that supply the majority of the profit.

Every visitor is a cost to the web site.  Even organic search visitors require the expenditure in time and money in creating the web site, maintenance, hosting, customer service and resources spent on achieving a higher search engine position.  Visitors through banner campaigns, affiliate networks and pay per click models have an obvious up front cost to aquire visitors.

You should use your web traffic tools to identify how many visitors are being captured by each source and place a cost on each of those visitors.  This is will be the Cost of Acquisition Per Visitor.

You should then calculate the amount of revenue generated by each source by analyzing the navigation paths to your conversion goals.  The relationship between the cost of each source and the revenue they generate will then give you an overall indication of which source provides the most valuable customers.

Similar pricing models can be developed based upon search terms, third party links, search engines and even browser agent details and screen resolutions. 

QUALITY MEASUREMENTS

When defining a quality measurement it must be defined against a specific target goal.  That goal could be a sign up form, a purchase, a subscription or just about anything that you believe adds value to your operation.  If you gain your revenue from banner displays then it may merely be how many page views per day are being generated.

These quality measurements become your primary KPI's.  They are how you define whether your web site operation is successful.  It is these that you will use to make decisions about what future changes you make to your web site content.

These quality KPI's also need to be measured over a time period.  If you alter your content then you need to record the KPI value for a period prior to the change and then record it for the same period subsequent to the change.  Only in this manner can you make an unbiased comparison.

You may find it helpful if you combine these quality KPI's into a dashboard format that can be easily viewed and understood.

RETURN ON INVESTMENT

The most important KPI measurement you can measure is your Return On Investment KPI.

An ROI can be  based upon visitor traffic, products, web pages, emails or any other online activity than can be reliably measured and calculated.  It is the ratio between the cost of generating the activity and the revenue generated by it.  You can start by creating an ROI for the entire web site and then drill down into specific activities on the site.

The easiest return on investment KPI to measure are those for online marketing campaigns.  CPM rates for banner advertising and pay per click costs for PPC campaigns can be easily recorded.  The conversion results for these campaigns can be captured  and an ROI ratio calculated to show the effectiveness of campaign material.  Similarly, email url responses can be recorded in traffic logs and final conversions recorded against the cost of producing the email campaign.

Multi-variate A/B testing using alternative versions of campaign material will highlight how particular material affects ROI ratios.  This could include variations in pay-per-click adverts, banners designs, 3rd party site locations and landing page designs.  The campaign material providing a negative return would need to be either abandoned or redesigned.  Marketing campaigns showing a profitable ROI ratio can still be improved upon using the evidence all of your ROI measurements are giving you.

HIGH QUALITY TRAFFIC

High quality traffic represents those visitors who indicate a much higher conversion rate in your measurements and significantly add to your return on investment.  Most web sites will follow the Pareto Principal which states that 80% of the revenue is generated from 20% of the visitors.

The goal is to retain this high quality traffic so that these visitors return to your site.  They have already made a significant commitment on your web site and so are far more willing to repeat that commitment.

A good proportion of your web site content, promotional material and marketing campaigns should be directed towards retaining this high value traffic.  This could include providing them regular updates of news, information and new product lines.  Retaining existing traffic is generally more effective than gaining new traffic. But do not swamp them with promotional material and do not show any sign of begging them to return.  They must return of their own free will because of the proposal you are offering them.

High quality traffic will be shown in your web site metrics by higher conversion rates, increased return on investments, lower bounce rates, more pages views and more time spent on your site.

LOW QUALITY TRAFFIC

Low quality web site traffic is essentially that visitor traffic that does not convert towards your target goals.  Though this visitor traffic may not currently be converting it still provides an opportunity for you to retain a relationship with them.  This traffic is essentially taking part by expanding the audience that is aware of your brand image.

Whilst the opportunity arises you should attempt to isolate this traffic by encouraging them to make a small commitment on your web site.  This may involve reading an article, signing up to a newsletter, responding to a poll, downloading a pdf document or bookmarking one of your pages.

You may measure low quality traffic in several ways.  Other than showing very low conversion rates, it will also show higher bounce rates, lower page views and less time spent on your site.  Since it is traffic that is not willing to make a high commitment to what you offer you have the choice of either ignoring it or attempting to retain its interest.  If you do add content to your web site to try and retain low quality traffic then you must ensure that this content does not weaken the content you already use for your high quality traffic.

TRAFFIC SEGREGATION

It is not possible to present a single marketing message that pleases everybody.  Each person is an individual with varying beliefs, values and cultural backgrounds that generate different desires.  Some people may be very austere with their money whilst others may believe that decoration or scientific evidence justifies additional expenditure.

Web traffic tools allow us to segregate data by location, hardware, software, time of day and search phrase.  This segregation represents the acitvities of real people from different places and using different equipment.  In itself it can be used to help steer our web site design towards resource intensive technologies or simpler less resource intensive web pages.  It can also indicate the type and style of language we should be resorting to.

To delve deeper into our visitors we need to engage them in answering on-site questions regarding their background and lifestyle.  What age and gender are they?  Are they fans of a particular style or fashion trend?  Are they optimistic about the future or pessimistic?  The nature of your web site will attract a particular market segment towards it.  But even within that sector you can attempt to segment the type of customers and determine what is driving their desire to access information or make a purchase.  You can develop these into meaningful Personas as described in the Behavior section of this site.

For each segregated market you wil need to present a different message and probably varying web content.  Communicate to each of these market segments in a manner that matches their desires and the way they view the world.  For example, a web page promoting the sale of heavily discounted clothing may not appeal to a teenager who is desires the latest fashion but may appeal to the mother who is cost conscious and seeks durable clothing for her child.

RECENCY

It is generally accepted that those visitors who visited your site most recently are the people who are most likely to visit again.  I have some doubts about this assumption because it does not take into account the behavior that would be seen by examining the simple laws of probability.  These in themselves would predict a similar pattern to that predicted by the concept of recency.  If recorded in chart format, both would create a chart with a bulge of traffic with a low recency and a long narrowing tail.  But I would accept that the majority of traffic to a web site is normally returning traffic not new traffic, unless specific campaigns have been designed to target new traffic.