There are world-class analytics tools available for free from both Yahoo and Google that will get you more data than God ever intended you to have. Yet, perhaps shockingly, a vast majority of decisions website owners make are based on faith rather than data.
It breaks my heart.
After wiping off my tears and meeting with many leaders of businesses large and small, I have come to the realization that the problem is not that we don't have enough data, nor even enough of the right type of data. The problem is we are obsessed about the data, and not about the business.
That has to change. Now. Especially if you want to take advantage of the most glorious channel on the face of the planet, the Internet. Here's the secret to creating truly data-driven organizations. Forget about the data. Well, at least at the beginning. Instead, focus on the business and answering existential questions about your business.
There are five things you need to identify to start your journey right. In this post I'll explain what these magical five things are, and share a concrete example with you. It is too easy to use an e-commerce website example, so I'll set myself a hard task and use as my example the website of a hypothetical non-profit, the Beta Robot Civil Rights Movement (BRCRM). If I can apply the framework to them, you can apply it in a jiffy to any other type of website.
Let's rock and roll. Here are the first five steps.
1: Identify Business Objectives.
Answer this stunningly simple question: Why does our website exist? A variation could be: What are the three most important priorities for our website?
Without a clearly defined list of business objectives you are doomed. Write down the answer to the above question, and ensure the objectives are DUMB (Doable. Understandable. Manageable. Beneficial.).
The responsibility of identifying business objectives belongs to the most senior leader in the organization responsible for the website (the person whose neck is on the line). There could be many objectives for the BRCRM but one primary business objective is to, for example, increase attendance at rallies and events.
2: Specify Goals.
Here's my definition: Goals are specific strategies you'll leverage to accomplish your business objectives. Since business objectives are usually strategic ("sell more stuff on the site" or "create happy customers"), goals help us drill down to something specific. Do X. Improve Y. Reduce Z.
The responsibility of identifying goals also rests primarily with senior leadership (in your company or at your client). Remember the people with their neck on the line? Yes, them. In our hypothetical example, BRCRM, the business objective was to increase attendance at rallies and the goal would be to increase sign-up on the website.
3: Distinguish Key Performance Indicators.
Your Web analytics tool, like Google Analytics, has a boatload of metrics in it (hurray!). A key performance indicator (KPI) is a metric that helps you understand how you are doing against your business goals (and objectives). Visits and Visitors are Metrics (as an example). Conversion Rate for an e-commerce site is a KPI. Makes sense?
Be very, very careful in identifying KPIs, because they are unique to every business (even between competitors) and a great way to fail quickly at this step is to do a shoddy job with steps Nos. 1 and 2 above. The responsibility of identifying KPIs primarily rests with the analyst (if you have one), or the marketer, with major amounts of input from senior leaders.
For BRCRM, given our goals, there are two objectives: First, number of visitors signing up for website event alerts and secondly, number of new individual memberships.
4: Set and Sweat Targets!
Targets are numerical values you have pre-determined as indicators of success or failure. You are going to love me for torturing you to come up with targets. I promise. This is a very hard step to complete, and yet it is super important. The challenge with having access to unlimited data in less than 60 minutes is that most of the time the person looking at all that data (and puking it all out!) has no idea what good or bad looks like.
It is your job to solve that problem. For your KPIs, what does success look like? What number equals failure? You identify that and when the report comes out the yardstick used to judge data, and freak out about it, has already been established.
The responsibility of identifying targets is shared by the analyst or marketer, and the finance team (and if no finance team then the boss). The analyst provides historical data and the finance team provides the forecasts, or just financial obligations, for the current quarter or year. Together both come up with targets for your web KPIs.
For BRCRM, as an example, we could identify: 1. KPI: event alerts. Target: 14,000 sign-ups per month. 2. KPI: individual memberships. Target: 4,800 per month.
5: Decide the Valuable Segments.
A segment contains a group of people identified by their source (where they came from), behavior (what they did on your website) or outcome (what goals were met).
A segment identifies people or their behavior that is important to your business. Either because you are spending money (paid search advertising, e-mail, etc.) or it causes you to have higher revenues (all people who convert). It helps focus where analysis of the data starts—what's important to investigate first to ensure data is in service to your business rather than the other way around.
If you want to learn more about this key step, check out this blog post: Web Analytics Segmentation: Do or Die, There Is No Try!
The responsibility of identifying the segments rests primarily with the analyst, with business guidance and prioritization by the senior leader.
For BRCRM, we have two groups of segments to focus on. First, acquisition. Since search is a great way to acquire new visitors, organic search is one segment. Since, for example, they are doing lots of e-mail campaigns, the visitors who get those e-mails are the second segment. 2. Behavior: Given our goals and KPIs, we have two segments to analyze and find insights from. People who sign up for event alerts and those who start new individual memberships.
Boom! With a wink and a nod you are ready, finally, to crack open your Web analytics tool and do something of business value with it.
OK, I'll concede the above five steps are not easy. They'll take a lot of back and forth inside the company and they'll require asking some tough questions related to the site's existence and business purpose. But when you are done with the five steps there will be no more data puking in your company, or decisions based on faith. There will be real analyses of the data, which will drive impactful business decisions that will increase your bottom line and create happy customers.
Avinash Kaushik is the author of Web Analytics 2.0 and is digital marketing evangelist for Google. Through his role, Avinash puts a common-sense framework around the often-frenetic world of Web analytics and combines that with the philosophy that investing in talented analysts is the key to long-term success.