Leave a comment
Get the GH Bookmarklet

Ask GH

I've been asked this a ton of times.  Further, I seem to be bombarded with tons of different potential tools.  I know that fundamentally, the technology underneath almost all analytics tools are the same, at least for straight up things people do on websites, but presentation and use cases vary.  When someone comes to me and asks for an analytics tool - how should I evaluate the business and the tool to get the best matches?  What should I look for in tools?

  • SE

    Sean Ellis

    about 7 years ago #

    I figure others may have a different view on this, but here's how I think about analytics tools. I'll start by answering from the negative perspective. What are the biggest issues with most analytics tools?

    1. They don't track something that I feel like I really need to know (like retention cohorts).
    2. They overwhelm me with information, while making it hard to find the truly important stuff.

    For me the most important metrics tend to center around marketing ROI, attribution and engagement. But your needs may be different than mine. You need to be clear about your goals before starting to evaluate analytics tools. For example, will you be doing lots of A/B or MV testing? If so, you'll probably want your testing tool to integrate with your analytics solution. On the surface level a test version might look like a winner, but when you track deep into a funnel, I often find that the winner is another version. If Optimizely didn't integrate with my analytics solution, I would probably be making a lot of incorrect decisions.

    Hope this helps get the conversation started.

    • SC

      Shana Carp

      about 7 years ago #

      For me personally, I find not integrating with my analytics solution necessary, mostly because I dislike the mathematical approach of most a/b/MV tools out there.

      I had at one point a script written for me that uses bayes's rule to find the right answer in an a/b test given my master metric. ( http://en.wikipedia.org/wiki/Bayes%27_rule )

      But that isn't necessarily true of my clients

  • JP

    Jason Peck

    over 4 years ago #

    I'll echo what Sean said re: being clear about your goals. A great thing to do is to start by making a list of the key behaviors you want to track and then prioritize them. Then you can begin to think about the "how" in terms of 1) technical setup (e.g. are you passing event data via an API or using "viewed XYZ page" as an indicator of the behavior/goal) and 2) how you're being presented with the data by the analytics tool.

    I'd also think about what actions you'll be taking based off of the data. E.g., are you looking to simply see the data and proactively seek it out? Or do you want some sort of alerts to be sent to you? This will at least help you start to ask questions of various vendors/tools to see what they can and can't do. Hope that helps!

  • RV

    Ravi Vyas

    over 4 years ago #

    Here is a checklist of sorts.

    - Is the tool sampling data or not
    - How far back does the tool allow you to look at? Some tools only allow you to segment data which is 30-90 days old.
    - Can you build custom dashboard for quick views or reports?
    - Does it have capabilities of pivoting on data ? (I love pivots, they give a lot of insights :smiley: )
    - Does it connect with other tools? (attribution or engagement tools for example)
    - How much data do they allow you to track. GA Events for e.g. can’t track a lot of event properties, other tools limit you to the number of events.

  • SC

    sri c

    over 4 years ago #

    In Avinash Kaushik's blog http://www.kaushik.net/avinash/sitemap/
    scroll down to this section - Tools, Tools, Tools!: "How to pick? What's the best out there? How can I be smart?" and you will find some blogposts on analytics tool selection.

  • SR

    shwaytaj r

    over 4 years ago #

    1) Do I understand what data is being represented in the first 3 minutes.
    2) Can it clearly explain what it does and does not measure. What are the 3 things I can measure and what are the ones I can't
    3) Can I do everything I want with this tool, or do I need others to make up for the void

  • MB

    Melinda Byerley

    about 7 years ago #

    Do you mean free or paid tools?

    • SC

      Shana Carp

      about 7 years ago #

      Either/or - I find even with Google Analytics, some people might need to be bumped to a paid tier.

SHARE
24
24