Tuesday, August 31, 2010

A journey into Web Analytics (Part III): Critical factors for success

Success or failure? Choose your way!
[It is time to resume the series I started in June. After presenting the business value of Web Analytics (Part I) and the some of the key challenges and obstacles lying on the path to successful Web analytics (Part II), let’s discuss what it takes to get a chance to reach Analytics nirvana]

“Tactics without strategy is the noise before defeat”
Web Analytics can start as an “underground” activity i.e. activity that is part of project or within a team. You can do some analysis here and there, making some recommendations or encouraging actions. It can develop on its own and grow larger in a certain extent but at some point – it will stagnate.

Don’t hope that you will be granted everything you need (i.e. money, resources, tools, responsibilities) “naturally” just because you do understand the value of Web analytics, because you are a analytics super-ninja and doing a good job. It doesn’t work like that, believe me! I have been there, I have done that and reality bit me, hard. Ouch!

If you want to turn this “activity” into a “culture”, if you want Web analytics to grow across the organisation, you will need to come with a strategy. And preferably a good one if you don’t want to see Web Analytics becoming your Waterloo.
Better have a good strategy

Critical factors for success
Ok, saying “you need to have a strategy” is easy to say – anyone can do that. But more practically, what does it mean? Your strategy - your battle plan -  has to address several key aspects – what Stéphane Hamel presents as the critical factors for success in his Online Analytics Maturity Model:
  • Management & governance
  • Objectives & Scope
  • Process & methodology
  • Resources & organization
  • Technology & tools
You will need to make sure that your strategy covers each of these factors. Now, keep in mind that addressing these will not guarantee success – but not covering these will certainly lead to failure. Well, that’s what I have heard and based on my own frustrations…er, experiences – I tend to believe it :-)

Getting top management support
Get manager support to leverage online analytics across the organization
I think that management factor is certainly one of the most important, at least it is the one you should address first. If you want Web Analytics practices to become part of your organization culture, getting top management buy-in and support is a must do!

Without it, it will not be possible to implement the necessary process & organization changes, you will not get the budgets you need for resources & technology and you may address futile objectives.

So try to identify what level of management you need to convince in order to get enough support to implement (and enforce) new processes, new organization and very important, to get money. Make sure you get real commitment - management will have to believe in the necessity of developing an online measurement culture. Your key sponsor will have to be ready to fight and die for your cause.

Don’t underestimate the task - getting management support is NOT that easy – it requires an appropriate “language”, to present complex concepts from a totally different perspective. It is very likely that what is soooo obvious for you will not be for them.

Scope & objectives
Once you get a general to lead your war, you will need to have objectives! What problems will you try to solve? How will online analytics contribute to your company bottom-line? Which area will be covered: internet sites, intranet sites, mobile, social media, other?

As said in part I, your ultimate objective should be to contribute to at least one of the three essential goals: increase customer satisfaction (customer first, remember?), increase revenues and decrease costs. But you will need to be a bit more precise than that.

You will need to break down your macro objectives in smaller and reachable objectives in a “SMART”(*) way. Tie these to your company raison d’être so these have a meaning for top management and executives. You will certainly end up with a long list, prioritization will be necessary. Where can you get most added-value? Are there any quick win? It will be a question of compromise.

Don’t be too ambitious - try not to address too much areas and objectives at the same time, if you don’t have the capabilities to do so. It is better to do few things fully rather than a bit of everything (that will bring no or very little value). Extend your scope and objectives as you get more “mature”.

Process & methodology
Process is required to make online analytics a systematic practice
In previous part, I explained that change management is one of the thoughest challenges (like in many other areas). How to make people change the way they work? How to make web analytics practice something systematic? The answer is you need the right processes and methodology that you can apply across the organisation on all online projects.

Many people tend to be allergic the term “new process” :-) - it often evokes time-consuming-useless-annoying tasks documented in a 100-pages document written by someone who will never do the work. Actually, A process can be very simple. Instead of reinventing the wheel, try to link Web analytics practices to existing processes. Keep it simple and flexible! Avoid big changes (it will not work) but rather start smoothly and then enrich your process, improve it on a regular basis (in true Kaizen way). You also can look for inspiration in proven processes like Six-Sigma or PDCA.

Don’t forget: a process is useless if no one is aware of it – so communicate it inside the organization using top management support.

Additionally, you need a methodology for turning business requirements & objectives into key performance indicators (KPI’s), metrics and reports. You can invent your own or, again, look for existing ones and adapt these to your context. I suggest having a look at the Nokia methodology – it is very simple and flexible, it has been more than a source of inspiration for me. Steve Jackson’s “Cult of Analytics” book also details a methodology to develop KPI’s

Resources & organizations
In Web Analytics, people are essential. Ideally, you need a good mix of varied competencies in technologies, analytics and business among others. It may be difficult to find all these in one single “champion” – so you may need to look for different persons, either internally or externally. Getting help from outside is definitely an option to consider – especially in the beginning – as it can bring valuable experience. If you have someone internally, take care of her/him – talented people are hard to find and very demanded (just have a look at this recent job trends chart).

But having the resources is one thing, defining the right organization is another one. For large organizations, I strongly believe in models like the “centralized decentralization” model (as Avinash Kaushik refers to it in his latest book) or the hub & spokes" model (as described by Eric T. Peterson for example).  The various skills are gathered in a central team that makes the bridge between the business & technology so the business people can focus on the thing they do best: business. The Analytics team will deal with all other aspects like what data to measure, how to collect it (tools, technology, and implementation) and it will teach how to use it.

Hub and Spokes model from Eric T. Peterson
The big question will quickly arise: where should Web Analytics sit in the organization? I raised the question previously and to be honest, I guess that the answer is that “it depends”. Wherever it sits, it should be as close as possible to the business. There is no point to have it at a level where recommendations are made too far away from decision makers, resulting in no or slow action.

Technology & tools
Web Analytics is like gardening - you need different tools
Last but not least, comes the technology i.e. the tools. Web Analytics is like gardening: you need different tools, each having a specific purpose. You can’t do gardening with just a shovel. Have you ever tried to cut hedges with a shovel (even if the most expensive one)?

It is the same with Web Analytics where you need different tools for different purposes: for measuring clickstream data, voice of the customer, competitive intelligence, social media, testing and others. Welcome to Web Analytics 2.0 and Avinash Kaushik’s multiplicity model! A model that definitely worth discovering.

But don’t get me wrong the number of tools is not important – it is what you do with them. It makes no sense to have a tool if there no use for it in your context – the one with most analytics tools doesn’t win anything. Based on your objectives and associated KPI’s, identify the type of data you needs and the required tools. Make sure to “integrate” any new tool in your existing framework – here again, adding these one by one will be easier than having to integrate them all in one go.

Know where you are and where you want to go
To get started with your online strategy, it is important to know where you are (your current situation) and where you want to go (the ideal / target situation). One way to do this is to use Stéphane Hamel’s Online analytics maturity assessment tool. Do a true and honest assessment of your current situation (or ask Stéphane to do it for you). It will help you to see where are the gaps (and how big they are) and what factors you need to address first in your own context.

I strongly believe that addressing these factors is essential for developing a successful online analytics culture on the long term – especially in large organizations. For sure it is easier said than done - building an online analytics culture is a long journey. This is the path I decided to take for the best and worst. Future will tell me how a good or bad choice it was.

I would be curious to hear your feedback and own experience regarding these aspects. Is it the right way to go or disillusion awaits me? Don’t hesitate to leave a comment.


(*) SMART = Specific, Measurable, Attainable, Realistic and Timely

(Next and last part: Web Analytics - a new profession?)

Related resources:
Related posts:

2 comments:

  1. Great article! I enjoyed all Parts I, II and III and the posts validates what I also experience with web analytics.

    It is indeed a long journey with lots of roadblocks and the battle to drive value out of is is endless. We need a lot of determination and passion in this industry. Very often other people in the organisation do not speak the same language as we do as web analytics specialists. It is crucial to translate what we want to do with analytics in layman terms.

    Creating a data driven culture is a long term process and this is very challenging.

    ReplyDelete
  2. @Joan: Thanks for commenting. Happy you appreciated this series. This is a long quest and sometimes it is really hard not to feel like Don Quichotte, fighting windmills :-)

    Another dangerous pitfall for Web analyst is to get stuck in reporting monkey tasks rather than true analysis tasks. Once you get in there, it is very difficult to step out and it eats too much valuable time.

    Cheers,

    Michael

    ReplyDelete

Feedback Form