Alright fellow data crunchers, you ready for the final post in our three-post series, "So You Call Yourself an Analyst?" I was really hoping to get this post up last week but instead I was putting the final touches on our new affiliate program (more on that very soon)! For those of you that missed Part 1 or Part 2 of this series, I suggest checking those out before reading this one.
In Part 2, we outlined ways to analyze the data for valuable insights. We walked through GA features, as well as strategies to help you prioritize your analysis efforts. We should all be sitting here with tons of great data just waiting to be used.
Assigning Value to the Data
So what do we do now? Buy ourselves a pumpkin spice latte, and call it a day? No, no, no. It's our job to take all of this data and figure out how much we are making (or not making) from our efforts. We can begin to create options based on data to help steer our companies down a more profitable path. By assigning value to the data you are now speaking a language everyone understands--money.
You need to explain the current situation in dollars. Too often analysts show up to meetings with great data but they explain it in weekly increases or decreases. This can be hard for people to conceptualize. Instead translate the data into actual dollars. I find the value per visit (VPV) metric to be one of the most overlooked calculations an analyst can use. Let's roll through an example;
Let's say we wanted to know if we should allocate more resources to the Learn SEO section. We would want to know how much each visit, on average, makes us. How do we calculate the value per visit? We pull how many people visited the section last month, and see how many conversions took place. We then multiply the # of conversions by $99 (you can also use your lifetime value if you want) and we get a monetary value the section brough in last month. We take that value and divide by number of visits to the Learn SEO section, and we have a practical data point to report. Looks like this;
.png)
Reporting good data with an attached value to your boss or client is only half the battle. The other part is making those numbers compelling. For example, let's say your marketing team needs to know what section of the site they should focus CRO efforts on. You may be inclined to throw that data in excel, but why not throw it in a pie chart? The data really pops, right?
When reporting the value of data you should lean on things like bar graphs, pie charts, and ranges of success. Go dramatic folks! You just spent days throwing back caffeinated drinks so you could find data they want to see. Help them see it better.
Last but not least...Take Action!
Now that we have the value nailed down, we can move onto my favorite part--taking action. Yup you read that right...it is your responsibility to hand the rest of the company with suggested "next steps" {mind explodes!}
Now I want you guys to be honest, how many of you have said something like this, "If we can increase traffic to this page by 10% we will increase signups by 2% and add X number of dollars to the bottomline?" Ugh, come on people! You have the data. You know how the traffic is getting to that page right now. You know what referrals are increasing, and which are dropping. You know what sections of the site are funneling to that page and which have no way for a visitor to access that page. Make actual suggestions when presenting the data!
You should be saying, "Here are my suggestions for increasing traffic to this page..." and list a number of referring sources for PR to work on, or give a handful of keywords for your SEO to target, etc. These are things you can suggest based on the numbers. You can have confidence that you are suggesting smart moves for your company.
That is what this was all for friends. As analysts we aren't just tasked with making sure everything looks okay, we are also expected to find ways to improve. If you aren't walking away from the data with action items you can get started on, you shouldn't be walking away from the data.
Well that about wraps it up for this series folks. I hope I have helped some of you reevaluate your current analytical process. I know it all seems like quite a bit of work, but trust me when I say time invested in analytics is well spent. Hopefully this series has given you a roadmap to follow, and at the very least motivated you to log in, check it all out, and start analyzing.
Want to start from the beginning and check out Parts 1 & 2? Here they are:
Part 1: Asking the Right Questions
Part 2: Analysis Redefined
As an analytics nerd in the making, I believe assigning a value to a visitor rather than to a visit is more practical cos not every visit leads to conversion and also a visitor can generate multiple visits. For e.g. lets say in the first three visits a person may be interested only in learning seo, so in those 3 visits the person will not sign up for pro membership. The same person may signup in say 4th visit. So when you calculate value per visitor, the focus is changed from volume to quality. So how the report may look like:
3450 visitors to Learn SEO last month
53 of those visitors resulted in Pro Signups (one visitor leads to one signup)
Calculate the total value of those visitors:
53 (conversions) * $99 (price of PRO) = $5247
Value per visitor = $5247/3450 = $1.52
For the same reason as outlined above the goal conversion rate should be percentage of site visitors which results in goal conversions and not percentage of site visits which results in goal conversions (as computed by google analytics). Another funny math used by Google Analytics is in the calculation of the overall site conversion rate. GA ridiculously add the conversion rate of each goal to determine the overall site conversion rate. So if you have setup say 10 goals and conversion rate of each goal is 10%, then GA will report your overall site conversion rate as 100%. Bingo! you have touched the pinnacle of CRO. Needless to say overall size conversion rate is meaningless cos you can't take any action on the basis of that.
When reporting data i lean on comparison bar graph, as data is pretty much useless if you can't compare it with site average. One other thing which i have found really useful is $index value of a page. Through this metric you can determine most profitable pages on your website i.e. the pages which were most frequently viewed prior to conversion or transaction. You can see this metric in your 'Top Content' report. Sort the $index column in descending order and then apply the 'weighted sort' to get the most actionable data. The pages that will pop up (in the page column) are the one you should concentrate on for your conversions. Happy Analyzing :)
Oh thanks for sharing. I like your approach, I think there is always more room for refinement in our formula based calculations. The awesome part is that you are even looking at it with this lens, it already puts you leaps and bounds ahead of most!
Assigning value to a visit is powerful. Most business owners think of visits as being moderate to low value, but when you measure which site sections drive business goals, and calculate a value per visit, you really can make the case for making improvements. Great post!
Great series Joanna, I really enjoyed it.
I would add one more component, cost data and accountability. It's always wonderful to show that revenue is up. But, it's even better to show that revenue is up by X, and that the cost of your salary and actions to increase revenue by x is actually x-y. With this bit of info in hand you can then politely end your presentation with, "Shall we talk about renewing my contract / increasing my bonus now". If revenue is up, but costs outweigh the gain, then you are better off doing nothing, assuming you can't find better solutions. Thanks again.
You make a really good point. That is kind of what I meant by "ranges" I think its our job to show them worse case and best case as well as variety of touchpoints in between. This allows us to give a realistic projection and then hopefully at the conclusion of our agreement we can show how well we did and renew that contract :)
You should go back to Part 1 and Part 2 and cross link all these for folks who come across those first. :)
Vic
Thanks Vic, I just remembered that as well. Going now :)
Another nice post - especially the ending. It might be the people up top who make the decisions in the end, but if an analyst can make the data accessible to them and make suggestions as to how things can be done then they're half way to making sure the right decisions are made. Or at least that the analyst's methods are used rather than someone else's.
SEO Services is an art that helps obscure websites lost in the search engine maze to achieve top rankings on search engines and bring profitable business.
Thanks for once again giving a numerically logical argument room to air round these parts. I am a huge fan of mathematical arguments and you have a way of articulating yourself that is both highly understandable and easy on the old synapses. I often feel that there is a huge room for more of this type of logic in all areas of our industry and welcome any move towards higher levels of sound mathematical reasoning. Can not wait for your next post!
Thanks so much! I am hoping to get up more analytic and formula focused posts in the future :)
I know I sound like a broken record er...um...a skipping CD but Joanna Lord, your posts ROCK!
You say it all when you say If you aren't walking away from the data with action items you can get started on, you shouldn't be walking away from the data.
Thanks for all the logical discussions you brought up here. More power!
Great guide, i cant wait for more lessons. Its all clear and simple as should be ;)
Thanks for the post, but theres a part I'm not understanding. How are you defining a visit to a section and then counting that towards the goal. If a visitor visits each of these sections (i.e. Homepage, then blog, then tools) and then converted, would you attribute this to each of them, or just that last section they were in before converting?
I haven't seen it done this way before and am more used to $ Index in Google Analytics which measures the power of every page viewed to lead to a conversion.
Thanks so much for this series. It has really helped clarify to my staff what sort of knowledge we need to get out of the information. People get overloaded quickly with all the info out there on websites.
Great tutorials. Thanks
I have a personal preference for presenting numbers in a spreadsheet. I recognise lots of other people need graphics. Does anybody have an easy excel to infographics approach?
Great post! In fact, great series {mind explodes!}
I plan on bringing some of your suggestions to the table and I look forward to future posts.
So true. This is why assigning goal values in Google Analytics is important (and easy.) Even if you don't have an exact value you can assign, such as if you're doing lead generation for a service that results in a different sales price for every customer, it still makes sense to assign an arbitrary value to each conversion so that you can at least get a relative value metric for your data.
Excellent post, Joanna. Most business owners simply don't go that far into assigning a value to each visit and hence give it an arbitrary number value devoid of factual evidence. As always, enjoyed SEOmoz's blog.
I suppose attributing value to visits is useful but you are still using a theory and none of that is 100% accurate. However, I see the point in using the methodology, it`s the closest thing we`ve got at the moment.