Analysis Ninjas: Leverage Custom Reports For Better Insights!

BloomHere is a key difference between Reporting Squirrels and Analysis Ninjas: The latter almost exclusively leverage custom reports (powered by advanced segmentation) and the former flirt with one standard report and then another and then other and in the best case scenario pull only half of their hair out.

There is nothing particularly wrong with the standard 19,000 reports in your web analytics tool. But they do represent the Vendor's best guess about what you should look at. Sometimes they even get it right.

Most of the time though your business is absolutely unique (even as it exists amongst hundreds of competitors) and it is absolutely important that you take your web analytics tool and mold it around you. The power that is given to you even in free tools like Yahoo! Web Analytics and Google Analytics can create a view of data that will help you find faster insights.

This post is inspired by a suggestion from Horia Neagu in reply to my tweet asking for blog post ideas. My thanks to Horia.

Horia's question was: How about a post entitled "10 Google Analytics Custom the Reports You Absolutely Must Set Up"?

I am not going to write about that, simply because the very idea that a report is custom means that there are probably no "ten standard custom reports" to set up.

I am going to share one recommendation and two ideas for making your own custom reports better.

This is a "teach a person to fish" type post. Sorry. :)

[UPDATE: If you want to download three advanced custom reports to do Page Efficiency Analysis, Visitor Acquisition Efficiency Analysis & Paid Search Performance Analysis please check out: 3 Awesome, Downloadable, Custom Web Analytics Reports]

No Goals, No Glory.

Here's a cliché: If you don't know where you are going, any road will take you there.

Nowhere is this more applicable than when it comes to trying to find insights from your data you can action.

You report your poor heart away, no one seems to be able to take anything you give and take action.

Often it is the case that you and I have not bothered to sit down with he HiPPO / the boss's boss and tried to understand what in the name of all that is holy and pure is our website trying to do!

What are the goals?

No custom report (or advanced segment, the life giving oxygen) was ever created without an answer to that question.

So ask that question. Get an answer before you go about your customization ways.

If your leaders / clients truly want wisdom from you they will answer the question. But it does happen sometimes that begging or throwing yourself at her/him does not elicit anything of value.

In those rarest of rare cases (after you have already submitted your resume to other companies that will cherish you for the golden child you are) try to figure these one or more of these three things out:

1. Where it the company currently spending money? Email marketing? Affiliate? Paid Search? Online PR?

And what's the biggest bucket?

Now go create your custom reports because if you can help the HiPPO's figure out how to reduce cost of acquisition they will love you more than you can imagine.

2. If possible, without violating HR policies, figure out what your boss's salary bonus is tied to.

Start doing analysis that will help your boss get a raise. A great goal to have, love and promotions likely.

3. Go visit your website, yes yes the one you have not used for a while. : ) Find out the single worst thing about it (should take you less than half hour of clicking around).

Now go look for data that will help you prove that the worst thing is the worst thing. Not a bad goal to have to fix what's completely broken, and people will listen.

Three good proxies if you have no goals to start with. Ideally you'll know what your Macro Conversion is so you'll start your analysis with a bang. Super ideal would be that you know both your macro and micro conversions!

Remember: No goals, no glory. Not for you. Not for your boss. Not for your company.

Custom Reporting Tip #1: Always, Always, Always Focus On The End To End.

One problem with standard web analytics reports is that the data you need is scattered all over the place, making it harder for you to find insights.

For example I am trying to figure out which pages stink and need fixing. In Yahoo! Web Analytics the standard report only shows Page Views and Average Time on Page. How much good will that do?

Or I want to figure out which sources of traffic I should make love to or divorce? The standard Google Analytics report spreads the data I need over four tabs.

Custom reports are good at solving this problem. Drag the dimension you need (traffic sources, landing pages etc) and analyze the data by choosing the metrics that tell the end to end story.

End to end has three pieces: Input. Onsite Activity. Outcome.

Here is my favorite, custom, traffic sources report:

google analytics custom traffic sources report

[Click on the image for a higher resolution version.]

By inputs I mean metrics that help you understand (in line with your goals) how well the "top of the funnel" (usually acquisition) is working.

visits new visitsIn my case that is measuring Visits (to know who is sending how much) and % of New Visits (to know who is sending how much that is of value to me – new visitors are very valuable in this case).

At a glance I have the information to start making some preliminary superficial judgments about performance.

By onsite activity I mean choosing metrics that help you understand the behavior of your visitors on your website (thus absolving your Acquisition team of any blame, perhaps!).

bounce rate average time on siteIn my case that is measuring Bounce Rate (not so fast Acquisition team, don't get me bad traffic! :)) and Average Time on Site (as a proxy of measuring if the landing pages are engaging visitors and as a proxy of how much each traffic bucket engages with the site).

Depending on my goals I would choose different onsite activity metrics for my custom report.

By outcomes I mean, well you don't need to know do you? You read this blog! I am all about outcomes, every day!!

goal conversions average valueIn my case the outcome metrics are Goal 1 (my macro conversion) and Average Value created for my website.

I could also have used $ Index or Per Visit Goal Value metrics if I were analyzing a non-ecommerce / content only website.

Remember Without a crisp articulation of outcomes every battle you fight will be lost, every day and you will live a very very unhappy life.

With these end to end metrics my custom report tells me stories that would otherwise take too long to piece together (or stories I might have missed completely).

One of the stunning realizations was just valuing Twitter traffic for example. (Click on the above report for a higher resolution report).

My twitter (social media) campaigns were doing exceptionally well. Lots of traffic (#3) overall, the second highest conversion rate (0.78%) and a Average Value that was not the best but rather sweet ($136 – which looks ever better when you compare the cost which is negligible).

Yet non focused traffic from twitter is not doing that well. 0.33% conversion and $39 average value. Pathetic.

I can now jump, like a na'vi, from row to row understand performance quickly and efficiently.

The power for a custom report that shows the end to end story.

It is so easy too.

For example here is the exact same custom report created in Yahoo! Web Analytics, just 30 seconds of drag and drop:

yahoo web analytics custom report

[Click on the image for a higher resolution version.]

Always go e2e. If you don't, you better have a good excuse!

Custom Reporting Tip #2: Create "Micro Eco-Systems".

I think I can honestly say that I have get to meet a single decision maker or a department or a company that has yet to tell me: "You know what the problem is Avinash? I don't get enough reports."

:)

We love spewing out data and pretty soon your company has 200 reports and I'll bet you $50 that not a single decision is actually based on data.

So fix it.

Create micro eco-systems.

What I mean are custom reports that do three things:

1. Reduce the number of reports (kill! kill! kill!) and yet coalesce information into one place.

2. Match metrics up with the audience that needs it. Personalize, personalize, personalize!

3. Force you, yes dear darling you, to talk to people and truly understand what motivates them (and then you create a report!).

Let's understand how to do this by looking at a real life example.

My goal is to create a "search ecosystem" report that collects different important pieces of data, for three different stakeholders, all into one place.

I do that by first understanding who all the stake holders are who'll need to use the data (let's hope!) and doing a simple stake holder interview to understand what their business goals are.

Now rather than spamming everyone with a report (that no one will find, I'll have a hard time version controlling, and other such pain), I'll just put it all together in one place (at least in Google Analytics due to a simple yet exceptional feature – tabs!!).

Here is a pictorial view of the process that I'll go through:

google analtyics custom micro ecosystem report

[Click on the image for a higher resolution version.]

My first "client" is the Acquisition team, they are responsible for spending the company's money wisely. They are measured on bringing new Visitors (potential customers) to the site.

I create a tab for them that shows Visits, New Visits, Bounce Rate and Average Time on Page (not site). I add the latter two because I want them to see their end to end view and I want them to realize they hold some level of responsibility for people not just coming, but also staying.

We have just one report, each day (God willing) they'll log in and see their own personalized sweet view of the data:

search traffic acquisition report

[Click on the image for a higher resolution version.]

But I am not done yet.

Next up is my HiPPO. Let's call him Paul.

Paul only cares about Revenue and all things connected to revenue. He does not care about any other metric. Nothing wrong with that.

Rather than creating another report for Paul I click on *Add Tab* simply do this:

add tab to a custom report

Create a person view for Paul. I throw in Visits (I have to give him some context and some Input metric) and Goal Conversion Rate (so he knows efficiency), Goal Value, Revenue, Shipping (because Paul is having us charge lots for shipping because he thinks of it as a profit center (!!), not great but remember I am personalizing).

Here's the resulting output:

search traffic hippo report

[Click on the image for a higher resolution version.]

You know what the result is?

Paul actually looks at the data every other day (and a bit deeper each Friday) because it does not contain crap. It contains just what he needs to do his job (find people to reward and find people to fire).

That is what you are going for. Taking people from data apathy to data usage.

[Oh yes, yes, I noticed Revenue and Shipping are zero in the above screenshot. I wish I could show you someone's real data! Not today. But you get what the report is trying to do.]

Finally there's Amy. Another key stake holder, but a tougher nut to crack. You see her bonus is tied only to Visits, a low bar if there ever was one.

So what do you do?

You can't over smart Amy, she is too smart for you (and probably a level or two higher).

You are going to lose her if you give her too much data.

You need to entice her to start using data, and restrain your smarts – you know you want to create a impressive 8 column report!

In this case I simply add a tab. It says Amy Chang (so she knows it is her very own personal report). It has Visits and Average Time on Site. I added Time on Site as a Outcome metric, just to keep up with my outcomes obsession.

search traffic amy report

[Click on the image for a higher resolution version.]

It is simple. It is effective.

It will get her to see just the data she wants (plus one more thing :)).

And here's the sweet part…. since this his an eco-system report perhaps she (and Paul as well) might see other pieces of data in other tabs and might be intrigued enough to ask you to add more metrics.

Then and only then and only at that time and only when you are asked (am I repeating myself?) add those metrics. It is vastly more likely that she (and Paul) will use the data.

There you go… one report that has all the data each stake holder needs personalized and customized.

No one is going to come to you and say: "hey want folder is my search report" or "I don't understand what all this data is saying" – it is personalized. And when you have to make changes, it is all in one place.

Well…

happy birthday

Such simple little things: tabs (or cup cakes :)).

Makes it so much easier for you to create a data democracy. And a bit sad that you can't do this with most paid web analytics tools today – yes you can create a custom report but the above report would be a one huge 13 column report that:

1. You would be able to read on your computer screen and

2. No one will understand because of spurious data unrelated to what they want and

3. Drive you into the arms of a multi tab excel spreadsheet (which will bring its own bucket of pain for you).

I hope all paid web analytics vendors will incorporate this feature, for the sake for our data democracy!

That's the story. Goals. End to End. Micro self contained eco-systems.

I was hoping to teach you how to fish, rather than just tell you which 10 reports to create. Regardless of the specifics of the reports and metrics above I hope you have learned a bit more as to how to think about approaching the issue and the important things to focus on.

Custom reports are a powerful way to take what looks overwhelming in web analtyics – REPORTS and DATA – and make it didapper. It is also a wonderful way to start the journey of your company, big or small, to start using data.

Ok now your turn.

Do you have custom reporting tips to share with us? What small or big thing you have done that really really worked for you? Have you tried end to end reports? How about micro eco-systems? What strategy completely failed? Got a custom report you think everyone in the world should be using?

Please share your stories / tips / bruises / successes.

I'll send the best one a copy of my new book Web Analytics 2.0.

UPDATE: It was hard to pick just one winner so a copy of the book goes out to SteveK (for advocating common sense!) and to Ali Shah (for emphasizing sharing of context). A bonus prize also goes to MGSeeley (for bringing a smile with his adorable analytics haiku!).

Comments

  1. 1

    Synchronicity – I have been planning a post for a while matching Horia’s question was: How about a post entitled “10 Google Analytics Custom Reports You Absolutely Must Set Up”?

    Of course, custom reports are intended to best fit the needs of an individual business, but it's inevitable that there will be some common custom reports that are setup which highlight what's missing in the standard reports for Google Analytics.

    I would be interested what others use. For me, the two most useful custom reports are:

    1. A custom landing page report using different engagement metrics from the standard report (and similar to your first custom traffic source above.

    2. Search performance by Geography drilling down to keywords in each major market, e.g. US, UK, Aus, etc

    I would then apply advance segments to these, esp paid and natural.

    I have seen the dashboards for different team members idea before and that's a good idea too – although difficult to define all the KPIs for a single person in a single report / view.

    For those to be useful, the team members need to know about the Advanced Segments concepts which I find in training, many still don't although there are increasing number of postings pushing this.

    Does anyone else have favourite / common custom reports?

  2. 2

    Avinash,

    Great post as always. Love the examples for both YWA and GA. I think my favorite custom report that I setup for PPC campaigns would have to be the day-parting/week-parting report in Google Analytics. I stumbled across it not long ago on Kate Morris' blog and have been using it ever since.

    My question for you is, what if you get a custom report with sampled data that is so far outta whack you can't do much with it? I.e. margin of error at +/- 50%, custom reports are extremely powerful but there is also a limit as to how granular you can go, obviously.

    Garry

  3. 3

    Hey Avinash – I find myself creating lots of Custom Reports, because every client's need is different and it's important to get the data that each client / each team needs in front of them.

    For me the two most common types of "stuff" that I do are:

    1. Creating Custom Reports that include Goal Completions (Not just conversion rates but actual Goal counts). Lots of folks don't like dealing with 3.89%, even though for folks like us, it's not a problem. They want to see whole numbers (15, 3, 1,409, 0), especially for Outcomes / Goal Counts

    2. Creating an Advanced Segment that fits my Custom Report. I have one report right now actually that includes all "Corporate" traffic (Email Marketing Campaigns + Social Media), so my Advanced Segment takes care of only displaying Email + Social Media, then my Custom Report has an Executive and a Marketing tab (each tab serving its unique purpose in life).

    Creating an Advanced Segment on top of a Custom Report is really cool because not only is it pin-pointed, segmented data, it's data "democracy" as you put it, as well.

  4. 4

    Again and again you save the day, Avinash. People won't appreciate/evangelize/look at things they don't understand. Your blog post today will help me make for a more understandable, personalized analytics world for the people that matter. Hello ecosystem reports!

  5. 5

    Very informative post, thanks Avinash.

    I love using custom reports to create a view that mimics my thought process. However, I find custom reports invaluable in that it forces you to ask the question, what part of the website do you care about? As you said, you cannot begin creating a custom report without first getting a better picture of what they need/want.

    The ability to use custom reports and then apply segments to them allows you a LOT of control over the data and what is being presented to the client.

    Is it odd that I find it more difficult to figure out what the HiPPO (who is the boss of my point of contact) wants then it is to actually create the custom report once we've established the goal?

    Who owns web analytics framework, part 2?

  6. 6

    Hi Avinash,

    As always, refreshing to see you have Yahoo Web Analytics screen-shots in there. Thanks! – and to reciprocate, let me provide a bit of stats (we are nerds after all, right?)

    The “custom report” is the 6th most used report in YWA in January 2010.

    Which just show that we (vendors) can and should never aspire to have every imaginable report included, as that just isn’t remotely possible. I actually did a study on this a while back, that showed that the “custom report” was THE most used report amongst enterprise users.

    http://visualrevenue.com/blog/2008/05/5-most-used-web-analytics-reports-usage.html

    anywho, great post ..and have a great week.

    Cheers d. :-)
    http://twitter.com/DennisMortensen

  7. 7

    […]
    There is also the issue of information? Once you’ve connected with other analysts and shared contact details, followed them on twitter you can see information being shared e.g Avinash’s recent tweet on “5 points on Google Privacy” or his blog on social media metrics. The dissemination of information via RSS and word-of-mouth is greater than ever before and I’m finding that at times I’m struggling to keep up with reading of blogs and tweets because there is simply too much.
    […]

  8. 8
    Ned Kumar says

    Hi Avinash-

    Once again, a practical, informative, and insightful post. Agree with you whole-heartedly on how custom-reports can save the day. I think the Herb Simon quote "A wealth of information creates a poverty of attention", is very true here. When you have too many "standard" reports, there is a lack of focus on next steps. On the contrary, a custom report created with a clear goal in mind forces you to evaluate the results and take action.

    Btw, I hope the use of custom-reports is not limited to Analysis Ninjas. I am hereby creating a new breed – the 'Reporting Ninjas' :-), Reporting Squirrels who are beginning to realize the "Force" of custom-reports.

  9. 9

    Dr. Chaffey: Both the reports in your comment are excellent.

    I hate to make recommendations :), but one of my favs is an acquisition dimension (Keyword, Email, Campaign, whatever relevant) and these metrics:

    Visits, Unique Visitors, Goal 1 Completions, Goal 1 Value, Goal 1 Conversion Rate.

    It allows me to look end to end, focus on a single goal and look at not just the conversion rate, which is what we always obsess about, but also the raw numbers.

    For context I might also throw in another important goal, then even more fun!

    Garry: With regards to the sampled data. The error margin is there to help you judge. If you get a error of +/- 50% I would use it with caution, if at all.

    My path typically is to reduce the time range I am looking at or the segmentation applied because both of those things would help reduce the amount of sampling applied and I would have the answer I am looking for.

    Of course if un-sampled data is needed then the API is a good way to take out all the data needed and then it can be analyzed.

    Joe: Thank you so much for adding a specific example of a custom report that leverages advanced segment. My favorite kind of report! :)

    Brian: Getting the the HiPPO to tell you what She wants is problem simply because often they honestly don't have a clue. And even if they did, in the rare cases when that happens, it tends to be a obvious that describes, at best. a macro conversion.

    With time this should change. But if for now you don't get an answer from them consider using one of the three ideas I had mentioned to get going. I am sure the HiPPO will see what you are doing is of value and in the near future come long for the ride!

    Ned: I love it: Reporting Ninjas! :)

    All kidding aside the honest to God truth is that we need Reporting Ninjas in large companies because of the way these companies are organized. And these RN's need to be very very good at creating hyper relevant goal oriented micro eco-systems.

    -Avinash.

  10. 10

    I'm diving into Custom Reports in Google Analytics and CoreMetrics with wild abandon after this great post. Having dabbled prior I wasn't sure how to best use this customization to my client's advantage but now I have some general ideas I can extrapolate. Thanks again!

  11. 11

    Catch up on the analytics blog every other weekend. My website doesnt generate enough data to be worth any analysis, but my company's portal definitely presents a great case study – especially since it deals with market data related to finance.

    Hope your comments are working fine. Didnt know that you had a separate comment DB

  12. 12

    Great ideas to help me do some "gatejumping". Thanks!

  13. 13

    I loved this post, Avinash. I found it ironic that I read this (including this quote "If you don’t know where you are going, any road will take you there.) right after another request in my inbox to build some reports for the sake of reports (i.e. no goals!).

    I've always found Advanced Segmentation to be the backbone of Google Analytics. All the 'out of the box' reports are there to support the drilldown done using segmentation. Knowing what you are looking for, instead of just looking, is the drum I, and probably any aspiring analyst ninja, beat on a daily basis.

    Don't just ask me for reports… Tell me what you are currently doing, what you are trying to do, and what your KPIs are.

    Thanks for explaining this so well!

    *Test comment submitted. I'll take that book please. ;)

  14. 14

    Avinash, fantastic post and very insightful. You "inspired" me to ask questions of my clients and from the responses we will all live happily ever after :)

  15. 15

    I always get a little giggly and excited like a small school girl when I spend time delving down deep into custom reports and then the person you created it for is really appreciative. For a grown man, this isn't something I should be admitting out loud.

    Seriously though, I think there is a better feeling – that is the feeling you get when you have trained someone and then they create a custom report for someone else. The the final person is appreciative and your protégé is really giggly like a school girl. Then you need to start doing it for the people who aren’t your protégé, but just normal people in the business.

    This solves your need of not really having enough time in the day to create custom reports because of all the fire fighting. Your staff are doing all the hard work (the little custom reports with easy things in them) and you are helping them out by training them, advising them on what to do with the reports and doing the more in depth stuff they don’t have the skill/time to do.

    This post made me the most happy though because it was my birthday yesterday and I took your little cup cakes as a deliberate gift to me :) – I know I’m self centred but I don’t care.

  16. 16

    Yes, I completely agree. Relevance is key. Targeting your reports and dashboard segments to the viewer/ report consumer is an extension of what we endeavor to practice with visitors landing on the very sites we, as web analysts, set out to measure and optimize.

    The custom reports you outlined are great because they breathe relevancy. Since report consumers are like squirrels collecting acorns for the winter hibernation, it is up to us Ninjas to come to the rescue with actionable insight.

    Therefore, I would add a list of metrics to the email message, explaining what each KPI means or is reflective of in terms of visitor experience… i.e. conversion, engagement, relevance, opportunity, etc. Followed by a gauge for the report consumer to take action when faced with sudden shifts in those metrics. And if those reports are analyzed before sending, a section for recommendations would be handy for the purpose of change-execution-management.

    A lot of managers have yet to grasp what all this stuff really means, so this would help cultivate the analytics culture.

    I learned all of the above from Web Analytics: An Hour a Day – which to me is like the bible, the old testament of my initiation into this industry. But alas, there is a new testament in town, and winning a copy of Web Analytics 2.0 would help me learn more about the advanced stuff, like social media and multi-channel campaign measurement (perhaps built from a custom report!). Of course, there is a good deed in purchasing it too, with proceeds going to charity. So either way, I guess I can't go wrong! :)

    Looking forward to the next post Avinash…

    Ali

  17. 17

    For a long time, I was the guy that all clients of a large web analytics solution provider had to go through to get a custom report approved, designed, and developed. While working in this capacity, I learned a great deal about what some of the largest online organizations consider "important" to their business and what it's worth to them to get that info. But at the same time, I learned that custom reports are only as good as the level of aptitude the report viewer has towards the information being displayed.

    Many clients spent thousands of dollars for my team to build out custom reports built for them, mostly for convenience of having data laid out perfectly to their liking. On some occasions, they would contract my team out of true business necessity of having analytics data manipulated to report on very specific business reporting purposes. In either case, I saw the pitfalls of making investments in customizing a software to accommodate reporting needs of the organization's analysts and report users.

    The pitfall being that users in our industry rarely stay in their reporting capacity for longer than a couple years at most, and new analysts coming in to use the reports may not have the aptitude to pick up things from where the previous analyst had left off. When this happens, the new analyst almost always puts in the request to revert the changes back to the standard reporting packages that come with well documented usage guides, and a support team from the analytics provider that is well versed in how to use the out-of-box solution.

    The point I'm trying to make is that custom reports can be great to fulfill your reporting needs, but make sure you exhaust all options of using the standard reporting capabilities of your analytics package. Yes I know Google Analytics is a free package <3, but you really need to consider the cost of your organization adopting a report that won't have documentation or a support staff (or user communities) that new analysts or others in the organization can lean on for reporting continuity. After all, what good is a report if no one knows how to use it?

  18. 18

    An analytic Haiku:

    Avinash Kaushik
    I await his every post
    like Spring's first flower

  19. 19

    Ali: Providing context is very important, thank you for adding that thought. Often numbers can't tell the whole story.

    I think Custom Reports can aid in creating a data democracy. Lots of people can get access to hyper relevant specific data tied to their business goals. They can then make tactical decisions.

    My experience is that the deeper strategic analysis will still be done by the Analysts or people for whom, say, 30% of the job is to analyze data. They will look at custom reports and drill down.

    Rest of the democracy might be able to make do with the reports themselves (if they are built with some context – like in my example I go down three levels for the first custom keyword report).

    Steve: An excellent point, one that I wholeheartedly agree with. Creating complicated custom reports (and let's throw custom metrics in there as well) to fit an organization's, or worse one HiPPO's, current world view can be a time and money sink. Especially if that is why the consultants got hired to do.

    My hope was to stress the use of custom reports to remove the dispersion of data that exists in standard reports with a view to, and this is key, tie to business goals.

    I know in the BI world we are very used to creating dashboards with drill downs in flash or fancy software to appease people's world view. That yields projects that go on for months and might not have ROI in terms of identifying actions.

    We should have a balance. The important part is to create a data democracy without spending a lot of time on making the UI look better or on trying to figure out how to have a bunch of dials and gauges on the dashboard!

    MG: I LOVE IT!!

    So absolutely kind of you to write such a sweet haiku (and so close to Valentine's day! :)).

    Thanks much!

    Avinash.

  20. 20

    Thanks a ton. I have been using custom reports for some time but always felt I was missing something. Guess what, I did not apply goals to the custom reports. Stupid of me. Thanks again.

  21. 21

    Very nicely explained. Thanks!

  22. 22

    Hi, this is my first comment, although I am regular reader of your writings/videos/book.

    The Custom report mentioned above has really brought insight into my knowledge of Web Analytics and I have successfully employed these in my projects.

    I have also added a report very similar to this i.e. Goal Conversion Rate Based on Visitor type i.e. new/returning/referer etc.

    Can you recommend me any of your blogpost about how to effectively measure "Goal Conversion with respect to Visitor Loyalty"

  23. 23

    Hi Avinash,

    Great post, I have just been merrily making some custom reports for various teams and it's great to see the different ways data can be cut.

    One question though – in GA you can only look at the top 4 goals (from the previous generation of GA) and not the other 16 that you can now create.

    Any idea when / if this will change because we have a site that has a lot of different goals that could do with being in GA custom reports.

  24. 24

    I find myself using a simple "Top Keywords by Sales" custom report for a small ecommerce site.

    Dimension: Keyword

    Metrics:
    Goal Completions
    Visits
    Goal Conversion rate
    Per Visit Goal Value
    Time on Site
    Pages/Visit
    Bounce Rate

    Quick and dirty for those HIPPOS who want to see actual sale figures tied to a key phrase. (as mentioned previously by Joe Teixeira)

    After the Keyword dimension, drill down to Source.

  25. 25

    Avinash,

    Once more,you nailed it. After being burnt several time, these days I am a pro at serving customized tabs to different strata of the client business.. However one metric I never miss out is benchmarking.That is something they( read Hippos) always ask for.. what is the industry/competitor doing? I encourage them to allow me to share data anonymously with Google so that we can access benchmarking in GA. Maybe it's not something common with more established businesses but I play in the small & medium business arena. We don't have too much access to trends.What are your thoughts on how to accommodate competitive/benchmarking data on the reporting tabs?

  26. 26

    Mohsin: Loyalty has several metrics in it, you'll find them in the custom reporting section under Metrics.

    For example for depth of visit you can use Page Depth or for Loyalty get Days Since Last Visit etc.

    Best way to measure success / "goal conversion" is to create an Advanced Segment with one of the above metrics and then apply it to the visitors report in GA.

    Rob: I can't comment on when more goals will be available in custom reports. I can tell you that I wish they were available yesterday – so there is an urgency. :)

    Bibi: The problem is that competitive data is not available in GA, to be chosen and shown on Custom Reports. For that you'll have to necessarily have to bring it in some place else (like excel spreadsheet etc). I don't like that, as you well imagine.

    But now I have started to use Annotations to show CI data. So I can manually type it in and then make sure I choose the right graph on top of my report and add the annotation and we are in business. I wish it was easier.

    Avinash.

  27. 27

    Avinash i love the idea of using the tabs for the stakeholders in the projects, i guess its one of those headslap moments :)

  28. 28

    Avinash,

    Incredible article. Like usual. Thank you for the insight, and inspiration.

    -Kevin

  29. 29

    Again, great article. One thing though really bothers me every day I open Google Analytics: Why can I include only the first 4 goals of the 16 possible into the custom report?

    That makes most custom reports useless for me, since a lot of important goals are goals 5+ on my website.

    I hope Google integrates those other goals into the custom report feature soon. Does anybody know whether this is planned?

  30. 30

    Ever since I listened to your Advanced Analytics class (a HubSpot webinar last year), I've been creating basic reports.

    Why?

    Because your definition of Bounce Rate cracked me up and woke me up! A high BR = "I came, I puked, I left" {I hope I quoted you correctly}.

    Although my traffic is not high, your post is still informative and I look forward to using some of your report examples in the future.

    To answer one of your questions, I don't have a great report to share. However, I would like to know if the following report can be created (I can't figure it out. Maybe it cannot be done because it needs data from my WordPress Blog but perhaps you know of a different way to get the end result).

    I want to know what portion of my new traffic, that stays long enough to read a post, are not the spammers that are dumped into my crud bucket. So I can learn what they want/need to read/learn about, where they live, the traffic source, etc.

    Hope that makes sense and if this can be done I'd love to learn how.

    Either way, thanks for all your valuable content Avinash!

  31. 31

    @jennifer i have had a client that needed this they were a content based website and knew that on average length of their articles was 600-900 mins

    "Apparently, an Average American adult reads at about 300 words per minute"

    So anyone who spent between 2-3 mins on a page had usually read the article, you can set this up within Advanced Segmentation, remember there is time on page & time on site…

    As for dropping out the spammers you can filter out using advanced segmentation 100% bounce rates, those who spend less than 5 seconds and even countries not relevant to your audience or language.

    You can now easily trigger a goal based on someone spending a certain time on page.

    Hope that helps :)

  32. 32

    @ David – Great data & I never thought about disregarding countries not relevant based upon language – simple but awesome point!

    Fun stuff and big thanks!!

  33. 33

    Avinash, how can I expand the number of results delivered in a scheduled custom report? I have a report scheduled to show visits, page views and a couple other metrics based on Service Provider. It only reports about 10 rows of data. I'd like to expand it to the maximum number of rows. What is the trick to doing this?

    Thanks,
    Brady

  34. 34

    Another way of saying it: If you do not know where you are going, all roads lead to failure.

  35. 35

    Jennifer: You can't report "those that stayed long enough to read the post", because "read the post" is hard to measure.

    What you can do is segment people who saw a post. You can segment out people who saw more than one post. You can also do something advanced and set a custom variable for each person who submits a comment (legit or spam) and analyze them appropriately by creating different buckets.

    Hope this helps a bit.

    David: Regarding your reply to Jennifer…

    If a person only reads one page and leaves a website then the web analytics tool has no idea how long someone spent on that page.

    Regardless of the web analytics tool you use the last page (or in case of bounce first page) registers the time as N/A or Zero. In as much Jennifer would not be able to do what you are suggesting. If she creates a segment with time on page then she will only report on people who saw the blog post page and at least one more page (as that one more page is required to compute time on the blog post page).

    More on this post about time on site/page:

    Standard Metrics Revisited: #4 : Time on Page & Time on Site

    -Avinash

  36. 36

    Brady: It is not possible to get more rows via email, I think simply because you could send out 1000 rows and just kill email! :)

    Using the GA API is perhaps the best way to go. You could pull the custom data, put it on a intranet site and people could click through on a email or a bookmarked url to see the latest data and get as many rows as they want.

    The GA API is here:

    http://code.google.com/apis/analytics/docs/

    Try the Data Feed Query Explorer to quickly create the query that you want for your data pull:

    http://code.google.com/apis/analytics/docs/gdata/gdataExplorer.html

    Have fun!

    Avinash.

  37. 37

    Hi Avinash,

    Would you recommend that Jennifer use iPerceptions to have her end-user tell her what they'd like to see as far as topics are concerned?

    @Jennifer – A good read, even if the survey route isn't the best route to proceed with.

    https://www.kaushik.net/avinash/2008/03/4q-the-best-online-survey-for-a-website-yours-free.html

  38. 38

    amazing as usual, many thanks for sharing your learnings!

    i have one small question about how to value the % new visits metric when applied to traffic sources. on one hand, it's true you want traffic sources that bring you lots of new visitors. on the other hand, we also want traffic sources that send us loyal visitors (even more if we take into account that organic visits, for instance, also take into account subsequent direct visits).

    so my question is: what's the right balance between acquiring new visitors, and getting loyal traffic? i know that "it depends", but when a source is very diverse, like organic, it is difficult to tell if 60% of new visitors is better than 40% new visitors. probably segmenting by keyword (brand, non brand for instance) is the way to go…. but i'd like to know what's your view on that.

    many thanks!

    cheers!

  39. 39

    Pere: You know my answer: It depends. :) And yes segmentation is absolutely fantastic to identify the insights.

    But let me add a bit more context.

    The report was specific to me. I really do want a massive part of my visits to be New Visits because I know that once they are here (perhaps after a couple of visits) I am going to push them into RSS. That way the content goes to them rather then them coming to it. I like that (and this obsession means this little blog now has 35,000 RSS subscribers).

    Now for other businesses that might be a silly strategy. On a newspaper site you want return visitors because they'll click on ads and you'll make money (RSS wont' work that well for making money sadly even if for the reader it is more convenient). On forums etc exactly the same thing. Recently for a hospital website I was hyper obsessed with trying to understand how to improve repeat visits (because people needed latest info etc).

    I want to separate the concept of repeat visits from repeat customers. I don't mix those two because there is an entirely different type of analysis and science that we all apply when trying to get someone who has already delivered an outcome (order, lead, sign up) to deliver additional outcomes.

    Hope this makes some sense!

    Avinash.

  40. 40

    I'm struggling with the first hurdle (goals), and would appreciate some advice.

    I deal with a site for a management/booking agency, and the site is largely comprised of profiles for the 400+ people they represent.

    The site was largely set up to pass information about these people (e.g. biogs, photos) to the people who hire them (the customers – think repeat visitors).

    A very large chunk of traffic also comes from fans of the people represented, fans who are googling their name (think new visitors).

    My question is this: what constitutes success? There aren't necessarily any indicators on the web that a profile has been used to complete a booking for the person represented, and there's no clear goal for what to do with the 'fans' traffic.

    Something else I'm struggling with: with 400+ profiles (there are around 3 pages per person), how can I aggregate typical behaviours across all of them? e.g. How can I see that the overall picture of how profile sub-pages (photos, biog) are doing to create some sort of average to benchmark against?

    Many thanks,

    B

  41. 41

    Ben: To answer your question to any degree of value a person would have to dig deeper into your business and why the website was created in the first place. But here are some ideas….

    If I was analyzing premierespeakers.com some straightforward goals are people who sign up for the "premier digital catalog", the number of Contact Us request, the number of Check Availability requests, I can use a unique phone number on the web and track the number of phone calls, clicks to the speaker's books etc (easily trackable links), the number of blog RSS subscribers added, and… I could keep going. Ok one more… how about using an on exit survey

    This blog post outline "micro conversions" for a number of businesses, perhaps it will spark your creativity:

    Measure Macro AND Micro Conversions

    Or perhaps this one about how I would set success measures for a Government site (very hard to do!!):

    Web Analytics Success Measurement For Government Websites

    With regards to your last question…. it is quite possible to do what you are looking for. My advice would be to please work with an authorized consultant for your web analytics tool. They are affordable and it's the best money you could spend (figure things out in a few hours rather than weeks :)).

    Avinash.

  42. 42

    Custom reports are a must.

    Most of our clients wants to see the status of their websites. I don't really use yahoo but Google Analytics has saved our butts more than one occasion. It brings a smile to the customer's face when you show their custom report especially when it includes every bit of information including their website.

Trackbacks

  1. […]
    Como creo que en muchas empresas podría pasar, en la mía ya existía una lista de 20 páginas con keywords y sus respectivas variaciones, producto de los pensamientos de los HiPPO’s (the highest paid person’s people opinion). Si bien, algunas de las palabras fueron acertadas, es decir, había mucha relación con nuestra oferta actual; la mayoría carecía de un volumen de búsqueda!!
    […]

  2. […]
    As I read back through this post before publishing it, I was struck by how far into the tactical mechanics of web analytics it is. The overwhelming majority of web analytics blog posts focus on step 5 and beyond — how to use the data to be an analysis ninja rather than a report monkey. Understanding the mechanics described here is a foundational step that will support all of that analysis work. I was incredibly fortunate, early in my web analytics career, to have an opportunity to run the migration from a log-based web analytics package to a tag-based solution.
    […]

  3. […] Analysis Ninjas: Leverage Custom Reports For Better Insights! (Avinash Kaushik) […]

  4. […]
    Como creo que en muchas empresas podría pasar, en la mía ya existía una lista de 20 páginas con keywords y sus respectivas variaciones, producto de los pensamientos de los HiPPO’s (the highest paid person’s people opinion). Si bien, algunas de las palabras fueron acertadas, es decir, había mucha relación con nuestra oferta actual; la mayoría carecía de un volumen de búsqueda!!
    […]

  5. […] –   Kaushik, Avinash, Analysis Ninjas: Leverage Custom Reports For Better Insights!, 10/02/2010 […]

  6. […]
    Here’s one tip: focus on outcomes. What are the business objects you’re trying to analyze? Pull metrics like conversions or revenue into your custom reports to get some idea of how your business is doing. Then look at the metrics that support outcomes.
    For more juicy information about custom reports you can read Avinash Kaushik’s excellent post about custom reports you should check it out. It will serve as a great roadmap for creating your own custom reports.
    […]

  7. […]
    We recommend that you build custom reports focused on goals and outcomes. You will find that the standard reports in Google Analytics either lack goal and ecommerce conversion data or they exist on another tab, which makes it difficult to see the appropriate metrics side-by-side.
    […]

  8. […]
    Here’s the crux of the issue – to get Google Analytics to track Social Engagement with more in-depth data, you need to do some fancy coding within your analytics script. This isn’t always within our skillset, or ability – so how do we solve this problem? I built a very simple, but easy to implement custom report that will show engagement data only from social sources. I basically took Avinash Kaushik’s Page Efficiency Analysis Report, revamped it for social networks, and included revenue and adjusted goal completion metrics. I call this the Social Engagement a la AK Report (click on the link to add it to your Analytics custom reports).
    […]

  9. […]
    Het is gemakkelijk om verdwaald te raken in het aanmaken van Custom Reports, maar wat maakt nu een fantastisch Custom Report? Om die vraag te beantwoorden, raad ik graag dit uitstekende artikel aan van niemand minder dan Avinash Kaushik.
    […]

  10. […]
    Instead of recreating the wheel every time, take about three minutes and learn how to create and use Custom Reports. If you're like me, once you've discovered how to use the feature, you'll want to get lost in making custom reports. But what constitutes a great custom report? For that question, I'll defer you to an excellent article from who else but Avinash Kaushik.
    […]

  11. […]
    Since Google is great at understanding the content of a website, users are bucketed into categories that allow advertisers to show relevant ads based their interests.  You can even see how users in different Interest Categories behave on your site directly in GA by viewing the standard reports under Audience > Interests or creating segments and custom reports.  (I won’t berate you about the importance of using custom reports and segments, but if you are spending most of your time in standard reports then I’ll let Avinash berate you).
    […]

  12. […]
    Avinash Kaushik: „Analysis Ninjas: Leverage Custom Reports For Better Insights!" https://www.kaushik.net/avinash/leverage-custom-web-analytics-reports-insights/
    […]

Add your Perspective

*