From ignorance to full control in only 6 steps

I am still laughing after writing this title, mainly because I don’t believe in “gurus” and their magic formulas. This post aims to give some tips that although they are generic and should always be contextualized, with these precautions can be very useful, indeed they are for me.

What do I mean with “from ignorance to full control in only 7 steps”?. I’m talking about the process that begins with a non-measured website to (and as Robert Kaplan said, if you can’t measure it you can’t manage it) and finishes with a full set of indicators that allows you to an efficient management of your online businesses.

Here we go with the following 6 steps (disbelieve in Gurus and magic formulas, including this one, are the first step to a successful project):

1- Understand your project: This seems pretty simple however which usually happens is that we have the idea in mind and we think we have a deep understanding of the project, but until we don’t formalize it (in paper) we don’t have a complete sense of what the project it. What other times happen is that you have an awesome “plan”, so strategic, but nobody knows what to do on a daily basis.
So the first step is as simple as writing down a paper with information as “what is” our project, “who” will be benefited with it, and “how” we will guarantee its existence, which means what things must occur to guarantee its subsistence.

2- Perform User experience management: Whow, this is Guru’s like term, but since they doesn’t exist we will demystified this term. User experience management, refers to practices aimed to improve the experience of a user with the points of contacts of your site. Regarding the current content, it is focused on understanding those interactions, identify issues and perform improvements.
To understands the above mention interactions we should begin by identifying the points of contact between your company website and your users or clients (both the point of contact and the contact process itself) and identify who that interactions are being performed.

Lets do it even easier by following some simple steps (why not? It worked to Henry Ford ;-) ):

a. Which actions can a user performs in your website?. For Example: view products, compare products, make a comment, get prices, compare prices, view shipping costs, contact the technical support, etc.

b. How your users perform those actions? For example: Contacting the technical support require navigating three pages then the users lands in the contact form where the user has to fill “n” fields, which are X simple to the users. Once the user clicks on “send the form” he receive an email saying that he will receive an answer whiting 48hours, that answer will satisfy in X% his needs.
Probably now you are thinking “how can I know all that information”? The answer is very simple, just go an answer that to your clients…yes, perform a survey. Don’t be afraid of performing surveys, users and clients are normally pleased of answering surveys (even more in Latin America). Normally my clients have the idea that users hate to answer surveys, but nothing further from the truth. If you do it with a correct frequency your users will love to tell you about their experiences with your company.

c. What things do you users wanted to perform in your website that are not available at the moment? For example: Your users may find useful calculating the shipping cost without leaving the first page.
If you still don’t get the point, just remember your experience with you cell phone provider. Which contact points do you have with them? And How is the experience during those contacts?
3- Measurement system design: Once we know “why” does our website exists, which are the contact points and how should be performed the interaction between your clients (users) and your website, it is important to formalize the processes of each contact point.

Lets try to explain it with the “technical support” example::

i. Page A, the user click on “Technical Support”.
ii. Page B, the user fill his Name, Client id, Title, and the question itself. Then the user accepts to send the email to the Technical Support.
iii. The user receives an email from your website saying that the “Technical Support” has received the question and it will be answered whiting 48 labor hours.

iv. The user receive (or not) the answer, which can (or not) satisfies the client information need.

v. If not, the client will (or wont) send another email to the “Technical Support” (online) or also contact the call center (which is another process).
After that the is recommendable to generates each process flowchart, because it shows in a very friendly way how the process are and makes easier to evaluate potential process improvements, or identify “restricted capacity resources”. Restricted capacity resources are persons or objects (computers, machines, phone systems, CRM systems, etc) that represent a constraint for the process flow.

4- Choose the Web Analytics Tool: Choosing the Web Analytics tool is a complex issue and should not be taken lightly. Here I propose a process that I found very useful to select the Web Analytics tool.

i. Determine the optimal technology: There are two main technologies on the market today, Logfile Analyzers and Page Tagging tools. None of those is better than the other itself, but depending on each project/company particular circumstances and needs.
1. Logfile analyzers: Should be installed in your own servers. It analyzes the log files generated by the server where the website is located. It requires an internal team for maintenance processes.
Even if we change the Analytics tool the information is always available in our servers and could be reprocessed once and over again.
It is really difficult to setup if we wanted to measure our website in a very detailed way.

2. Page tagging: This technology requires that you add a script (tag) in each page of your site. Those tags are very easy to customize making very simple to generate very detailed information and tailored information. This technology doesn’t require an ad-hoc team for maintenance processes. The measured information is hosted in the provider servers, so as soon as we change the provider we lose the past information and should start measuring the traffic from the beginning.

Based on the above, a bank may probably prefer a Log file Analyzer, even when they have to incur in a maintenance team cost, because adding the tag (from the page tagging tool) in some pages could represent a security issue.
On the other hand for an ecommerce website would be vital having as much information as they can, available for their users behavior understanding. It would be also undesirable charging their structure with a dedicated team for the Web Analytics tool maintenance that will for sure reduce its ROI (or ROCI). For the mentioned reason, an ecommerce site could prefer a Page Tagging solution (as a matter of fact, there are too many variables to be analyzed before understand which technology is the best for some particular Project, but I think it helps to understand the mental process for the decision making process).

ii. Web Analytics Tool sifting: DeIn this stage we must drop those tools that are not able to measure the above mentioned processes and identify the restricted capacity resources (Analytics 2.0 theory).

iii. Short list: From the remaining list we must identify those with the following profile.
1. Affordable cost, don’t forget the Avinash’s 90/10 rule.

2. Good reputation: Long time providing a efficient service without interruptions (average uptime).

3. Trusted Technical Support: Proven technical support acceptable answer time, and available in your (or your team) language. Misunderstandings and/or misinterpretations with technical stuff could cause big information troubles.

4. Usability: From the remaining tools, I suggest to select the one that it is friendlier for your team. If you have a professional Analytics team, then they would welcome a very flexible one, which gives them freedom of analysis, while if you have a non professional team, they will prefer something simple even when it means less flexible.

5- Implement: In order to achieve an efficient implementation its crucial being focused on the website processes and in correctly setting the Web Analytics tools to measure them in an effective way. After that its convenient perform QA into a testing environment (you can generate the action and finding out if the tool is measuring it correctly).

6- Identify the most relevant metrics: Once the Web Analytics tool is working properly, it is important to design the required dashboards. The dashboards provide each person within the organization with a set of metrics that help them make efficient decisions in a minimum time period.
For example, some person within the organization may be interested in a particular (from the above mentioned) process like “Customer Attention department” may wanted to know if their client online experience is good or bad, because the cost of an online client attention center is (in my experience) ten times lower than a regular call center.
I hope you find this information useful. Anyways, remember that each company it’s a completely particular world. Don’t try to adapt your company to the “best practices” but “best practices” to your company.


  1. Here we go with the following 6 steps (disbelieve in Gurus and magic formulas, including this one, are the first step to a successful project):

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.