Successful Digital Analytics Project Workflow

Successful Digital Analytics Project Workflow

How digital analytics people should get involved in a digital project ?

For every analytics project, every new digital campaign, new launch,… I usually go through those steps or some of them without noticing anymore. Recently Adobe released a new whitepaper “How to create a data-driven dynasty“, amoung other stuff this document was aiming to provide digital analytics practitioners best practices for a successful digital analytics program management. I found this chart below very useful and a keeper.

The chart “Analytics project workflow [copyright Adobe]” below is illustrating a successful workflow to an analytics project – assuming that every digital project is a also a analytics project :

From an agency perspective and from a client perspective as well, this digital analytics project workflow seems pretty ideal to make sure that analytics is always part of the equation and avoid the common pitfall where is analytics is involved after the launch – and most of the time it’s too late to get things right and meet the business owner expectations:

  1. Plan

  2. This step covers a critical processus when the business owner meet the analytics team to explain the campaign, launch… expectations from a business perspective. The analytics team will then translate the business requirements into a technical tagging guide for the IT team. This document will cover the basic tagging which are most of the time already implemented by default and especially the specific tagging related to the specific project mentionned (e.g. a microsite launch which main goal is its social interest to users: the main traffic metrics will be measured as well as the volume of login through social channels, the volume of shares…).

  3. Implement

  4. As simply as the title is saying, this step covers when the IT Team get clearly involved and implement the tagging. Depending on your ressources, the IT team maybe trained to the analytics tool you are using or not, as long as the analytics team is sensible to that, the implementation will go smoothly.

  5. Validate

  6. As for for every technical implementation, this part covers the testing on an testing environment before going LIVE. I’ll usually use tool such as httpfox and look at QueryString details to make sure that every thing is running smoothly, that the tags are fired when they should be, that the pageName is correct… this works with Google Analytis & Adobe SiteCatalyst.

  7. Launch

  8. Review

  9. Lastly, after reviewing the implementation live, making sure that you are capturing every data points that you need according to the business requirements, will come the time for reporting, analysis and possibly enhancements.
    This part is at last the one the business owner is awaiting for, it will allow him to know if and in which extend his campaign, launch… is successful, which channel driver is performing better, what actions are the users performing, what kind of improvements could be done… The analytics team will be the guarantee of data-driven decisions making.

What about you, does this analytics workflow sound good to you ? what kind of issues are you confronted in this process?

As usual thanks for reading me so far, if you liked this post, please spread the love…

2013 Digital Analytics Roadmap

2013 Digital Analytics Roadmap

Scroll down for a larger version
Analytics Roadmap
Happy New Year everyone!
Let’s kick off this year with a list of wishful thinking and future articles, I certainly hope I will be able to work deeply on all those subjects and more but I do realize that not everything will be achievable !
I am sharing this list as I intend to write about all those topics this year, so stay tuned if one those items – segmentation, landing page optimization, cross-platform analytics, campaign performance measurement, site search… – is a previous, present or future topic of interest.




  1. Segmentation based on Personae

    In a recent article, When Analytics is King, Segmentation and Targeting are Queens, I was giving my opinion on how important segmentation is to do a better analytics job. There is various ways of doing segmentation and one of the efficient one is to based your audience segmentation on personae. Personae concept goes beyond demographic segmentation, I will soon dive deeper into this subject in the meantime, this article helps to understand the usage of personae in digital marketing.

  2. Landing page optimization : almost there…

    This topic is very large, my objective here is to finish Tim Ash’ book: Landing Page Optimization: The Definitive Guide to Testing and Tuning for Conversions, this book covers how to prepare all types of content for testing, how to interpret results, recognize the seven common design mistakes, and much more.(Amazon quote). I start reading it few weeks ago, it’s a highly valuable book, can’t wait to share with you a full review of it.




  3. Attribution Modeling & Testing : How to improve campaign allocation performance measurement ?

    Hopefully, performance improvement is a journey, not a destination. When it comes to digital marketing and campaign budget allocation, the equation gets more and more complex as the channels are numerous (Search, Display, Facebook etc.) and understanding each channel contribution to your performance (revenue, micro conversions, user experience…) is a tough job. Attribution models and testing is a step in this journey.

  4. SiteSearch Optimization


    This hilarious video of Google Analytics reminds me, how important site search is, I remember reading in a usability study that approx. 30% of your visitors will ignore your navigation, content links and hero images and use ONLY your internal search tool to look for what they are searching.

    Well, I do hate common views however this one worth taking the time to dig a little deeper and capture the main metrics to help your visitors find what they are looking for in your website in a efficient way.

  5. Cross-Platform Analytics

    A few articles this month claims : 2013 is the year of mobile analytics or 2013 is still not the year of mobile analytics, 2013 is the year of … Well, i don’t really care as the platform of today is certainly the old man of tomorrow’s platform. So I want 2013 to be the year of cross platform analytics : mobile, tablets, desktop, Tv… Whatever ! Where your clients/prospects are, you want to be looking at!
    About that, I wish to use and write about and compare those unified audience cross-platform audience measurement tools :
    Media Metrix® Multi-Platform from comScore
    Universal Analytics from Google Analytics

  6. Participate as much as possible to Analysis Exchange

    I joined the Analysis Exchange as a mentor last year, I didn’t have the chance to participate yet, well I hope that 2013 will be the year of my contribution. This exchange is a pretty exiting project run by Web Analytics Demystified to increase practice of web analytics by providing free web analytics consulting to non-profits and NGOs around the world.

Thanks for reading so far ! If you liked this article please spread the love…

5 Good Reasons to have a Business Intelligence Analytics dashboard

5 Good Reasons to have a Business Intelligence Analytics dashboard

As shared previously in another article about Big Data Analytics, we’ve got tons of information available for us marketing manager / analyst… to take in account to make good business decisions.
Thus we need to be selective in the data we look at, each data should have a purpose, even if we have access to quantity lets aim for quality. A business intelligence analytics dashboard is one step towards this objective: consider qualitative, selective and intelligent data to make easier data-driven decision.

As for a car where the dashboard is the part of a car which contains some of the controls used for driving and the devices for measuring speed and distance, a dashboard from a business point of view is a document showing in a single view the most important KPI (key performance indicator metrics) that need to be measured and monitored to drive your business. A dashboard is not only displaying some of your metrics, the challenge is to display them in a intelligent way : telling a story with your metrics that will help making data driven business decision looking backward and forward.

I am lucky enough today in my new role at MRM Worldwide to work with companies that consider a business intelligence dashboard as a must-have or companies getting to it and part of my job is to help them to build it efficiently.
Let’s deep dive here : Why would your business need a BI dashboard ?

1. BI Analytics Dashboard: What are we talking about?

A dashboard is kind of a reporting variety. A dashboard is a web based tool that deliver selective and relevant KPIs in real time to targeted user. Those KPIs have to be displayed in a way simplistic way – KISS, keep it simple, s… – that include visual guidance as gauge, traffic lights… and the guidance should also come from the ability to intelligently analyze historical and real-time data to model the future.

To elaborate a little more on my last item, each chart/graph should be accompanied with a title such as “What question this chart is answering to”, a platform that send plain English emails to give some alerts or status update and directions in the actions to be taken, a platform that has to be explained and built with the users.

2. It’s 100% tied to your business objectives

That’s the spirit of it.
You don’t really need all the metrics that your analytics and else platforms are allowing you to see. For instance, when every morning I am looking at my blog web analytics reports and I just want to answer few questions : are my loyal readers still there, what about the new ones ? new comments, new shares, how far did they read my articles, did they subscribe, when and where did they drop ? (this is a very ego-killing job but that’s part of the game).
So I do use Google Analytics ability to set my custom dashboard, however I still need to open all my others tools (Clicktale, Feedburner, Bitly…) to get the whole picture and evaluate it versus the expected outcomes I had in mind.

I’m not running my blog as a company, however if I was, using a business intelligence analytics dashboard will allow “my” company to automate comparison with real metrics VS expected outcomes or target and help me increase my operational efficiency. A marketing dashboard offers you the possibility, amoung others stuff, to follow your completion rate towards your company goals and act accordingly. For instance, knowing your completion rate you could use some predictive analytics (i.e. forecasting) to adjust either your goal or your campaign effort (spend, segment…) in order to meet your goals. A dashboard is not going to do the marketer job however all historical data could help to set up some alerts in case your campaign is heading in the wrong direction : run a SEO audit, spend more on this channel, test a new promotional call-to-action…

As a real life example, have a look at the data below from TNS Digital Life : The role of the consumer voice. It does make sense to compare your own site data to TNS data from a benchmark point of view, you’ll easily see then where are your strength and weaknesses:

3. It’s customizable & (B) Intelligent!

It has to be customized to your audience or there is no point having a dashboard.
Your CEO will not look at the same KPI as your Marketing Director, Manager or Regional Director or CMO… however everyone need to see the Big Picture, working in silos is not efficient and it’s frustrating (just my point of view). As you may be using segmentation and targeting for your campaign online or your website content – the same rule apply here : adapt the content to your audience. And as it’s web-based tool, customization could be endless thanks to segmentation, behavioral targeting, widgets for instance and mobility : a dashboard as to be mobile friendly as well. That’s for the customization part. What about “Intelligent”.
What is Business Intelligence ? What is BI when it comes to digital marketing?
From my point of view, it’s when you reach the point of telling a story through your data. To tell a story, your data have to be meaningful :

Well, I think the visual speak for himself, you should fell more guided with the right-hand side version of the visit’s metric. One the right version, you will be displaying the data but also evaluating it to your goal, you will use predictive analytics to measure your chance of success and finally use a green light to let your user know that it’s on track.

According to Gartner, Business intelligence (BI) is an umbrella term that includes the applications, infrastructure and tools, and best practices that enable access to and analysis of information to improve and optimize decisions and performance.

 

4. It’s a time savior.

Better use your time to analyse, look for pattern and trends, take decison than building the tool to help you make those decisions.
As a marketing manager, I guess remembering your Facebook Insight, Twitter, Google Analytics, Google Webmaster tools, Google Adwords, Bing AdCenter, Doubleclick, Customer Feedback tool, TweetReach, Mailchimp, Clicktale, Salesforce… and so on accounts, logging to each of them pulling data in your excel file and wrap them up together to get an overall view is not your priority, right? Even though Google made a great job of integrating each of his service into Google Analytics, we are not all Google-dependant some use multiple analytics tools : SiteCatalyst, Webtrends… and your CRM, Emailing, Feedback… platforms are not Google either.
So time is money and if you want to be able to catch the whole story & be close as possible to your customer journey, you may want to consider a dashboard.

5. Because it’s fancylicious 😉

Firstly you would be able to sound very intelligent, I can’t help thinking that Business Intelligence is a very cool word.
Anyway, working for data don’t have to be boring some even that it’s sexy: … I do LOVE excel spreadsheets and plain and simple table with black figures on a white background – always very efficient but I love infographics as well, so why not having both when it’s possible and relevant. To perform efficient data analysis, data visualization matters a lot. And there is a lot of magnificent dashboard that maybe will make you love data, find here my pearltrees about data visualization.

Thanks for reading! If you liked this post, spread the love…

Sangomar 2012, a lead generation oriented website!

Sangomar2012.com is a website I created for a friend. This website was created on Senegal presidential election occasion to engage the senegalese youth to vote!

I am certainly not a developer, but I really cannot hold on looking into the code, trying to code some stuff myself particularly as WordPress is really easy-going and flexible and also because my companion is web developer… so if I mess things up – as it often happens he could say – he can fix it!
Anyway, don’t want to bother you with my personal life, I just wanted to introduce my last “freelancer” website project Sangomar2012!
One of the reason why, I’ve been kind on lazy those last weeks on the blog, I was doing a website WordPress based for a friend. My friend choosed a wordpress compatible template + his domain and I took care of setting up the website for him. In few words, it was for me the occasion to build from scratch a website with WordPress (except for my blog).

      – From my friend/client point of view, this website was meant to promote and engage young senegalese to vote during the last election. As you may know, we just get the final results a few days ago.
      – From a webmarketer point of view, the main goal of this website is lead generation that’s why I put my effort in making the newsletter subscription form prominent and redundant but not too overwhelming. I used both Google Feedburner email & RSS feed functionalities and also a simple php based form.









I tried to gather the best WP plugins for his needs and also for my curiosity, here is the list of plugin I used to implement the functionalities needed according to the website main objective.

  1. Google Analytics for WordPress, impossible for me not to… I’m too curious
  2. Really simple Facebook Twitter share buttons
  3. HeadSpace2, for SEO purpose to customize page title, description…
  4. SEO Smart Links, also for SEO pupose a great plugin to promote internal links
  5. Simple Google Sitemap XML
  6. Google Feedburner

Very few compared to the list of plugins I’m using for this blog ! This website is my very fisrt one on my own, please feel free to comment, to criticise as long at it can help me to improve…

Allow me to make an aparthe, I’m pretty happy that I spend some time on this website as besides being one of my friend website, this website echoed my personal opinion about why vote is important!

Tracking checkout conversion rate with Google Analytics or Omniture Site Catalyst

About conversion rate, I would like in this article to dig a little deeper into checkout conversion rate and how to measure it. Currently, I’m doing an AB Test to compare the performance between a classic step-to-step checkout with a less classic “accordion checkout” (not to be confonded with one-page checkout).
Few things to define before talking about tracking with either Google Analytics or Omniture Site Catalyst.

1. Conversion rate

Well, this design should do the trick

Conversion Rate illustration
Illustration from the great great Conversion Rate Experts blog

Our subject being checkout, we will assume that the website is an ecommerce one and as a consequence the main goal/action we want the user to take here is “placed an order”.

2. AB Testing

Also here i think an illustration will express it better than words

Few words about the AB testing subject here.








3. Checkout ergonomy and design

Just want to highlight the difference between : accordion checkout, classic step-to-step checkout and one page checkout. The checkout being the step just after the basket, not to be considered lazy but once again examples is better than words:

  1. Accordion checkout: Following the principle of an accordion, this kind of design hide and show the step following the user progression without leaving the page, it’s a “vertical” design using Ajax most of the time. When the user is taken to the checkout the first section is open and he can see the titles of the following sections just below.
  2. Classic checkout: The classic checkout is more a “horizontal” design, each step of the checkout = one dedicated page.
  3. Single Page Checkout: This kind of checkout design can be horizontal or vertical, the principle here is having everything on the same page and every fields open, better option for short checkout process > everything is visible at a glance.

Context being clear now, lets get quickly to measurement! First of all, most of the checkout being in multiple steps – whatever the kind of design you choose – I would advise to measure 3 things:

  1. Checkout conversion rate = sales / number of carts initiated
  2. Overall conversion rate = sales / unique visitors
  3. Fall out step by step = % of visitors who drop on each step
  4. The 2 first performance indicators can show you trend and the 2nd expecially allow you to compare your rate to market conversion rate: knowing that the formulas can depend but either way the 2nd formula is supposed to be the one the market use and communicate about.

    The fall out indicator is the one you should/could take more time to analyse and set up on your webanalytics tool.

    With Google Analytics

    You first have to set up your goal. In our case, you goal is the last page of the checkout, usually the Thank You page then you have to set up the funnel which is each page/section of your checkout process. This is useful only if your checkout is a multiple page checkout : new step = new page. For one page checkout or accordion checkout, I would implement events tracking to get info about the steps within the page, but I will dig into this in a later article. Here is what you should get from GA:

    Why do I love Google Analytics: because it’s flexible, any webmarketer can do this without involving development team!















    With SiteCatalyst

    You will first have to set the “pagename” in the tracking page properties. But most of the time, this should have been done when implementing Omniture SiteCatalyst the first time. With this, it’s also really easy and flexible as you just have to drag and drop your pagename into the “Fall out report” to build your report!
    Here is what you will get:
    What I do love about SiteCatalyst fallout report, is that you may need developement team help but you will be able to track PAGE and SECTION in a page so any kind of checkout design can be tracked!








    Well, tracking is the first step of optimisation: go this article to learn more about few tips to enhance your checkout flow and decrease your checkout abandon rate.