- Advanced Google Analytics Features
- Google Tag Manager for e-Commerce Stores
- Google Analytics Event Tracking
- Google Analytics Multi-Channel Funnel
- Google Analytics Spam – Don’t Be Fooled: Filter Bogus Traffic in Google Analytics
- Google Analytics Acquisition and Attribution
- 13 Google Analytics Tips & Tricks for Better Data
In this article we cover Google Analytics multi-channel funnel functionality. In our previous post we had a look at how to set up sales funnel tracking and visualization in Google Analytics. We learned how the funnel functions and what it can show us. Using these insights we can improve a great deal on our site.
We also know what the channels in Google Analytics are – the paths which our visitors take to land on the website. Funnel visualizations show us which online channels serve us best, and which funnels work for one customer segment and not others.
Now we will turn to the Google Analytics multi-channel funnel report. You can find it under the Conversions section of Reports.
Google Analytics multi-channel funnels help us to accurately attribute conversions to each channel and put values on them. From this, we can determine how long it took from the user’s first visit to their eventual conversion and what steps they took in between.
It’s the most valuable information a conversion optimizer can have.
The Multi-Channel Funnel report consists of the following reports:
- Assisted Conversions
- Top Conversion Path
- Time Lag
- Path Length
At the top of the Overview report, you’ll see the familiar graphicon that shows the time distribution of the conversions. What is more interesting to us is the Multi-Channel Conversion Visualizer.
The Visualizer graphically shows the number of conversions contributed by each of the channels in the form of circles. The size of the circle indicates the proportion of the conversions and the areas of intersection between the circles indicate the conversions that multiple channels contributed to.
Using the Google Demo Account example, we can see the visualizer for the transactions (aka. macro conversions).
According to this, the greatest single contributor to conversions for the Google store is direct traffic. It accounts for a total of 70.21% of all transactions on store site, either by being the last click before the transaction or being in the path leading to conversion. The second most important channel here is referral traffic, with 33.90% conversions attributed to it. We can see the direct and referral traffic overlap 19.5% of all conversions.
However, as this is only an approximation, we can only conjecture at the exact path the visitors took. From just this visualization we can’t come to any conclusions; it only helps us visualize the contributions.
To get really useful data, we must visit the other sections of the report – the Assisted Conversions and Conversion Path reports.
The Assisted Conversion report shows the specific number of conversions attributed to the channel, based on the last click attribution model. The table also contains the approximate value of the assisted conversions for the channel.
The most important piece of data in this report is the value in the last column of the table. This number is the proportion of total assisted to total direct conversions contributed by the channel. The higher the number in this column, the more the channel assisted conversions.
From this table, we can see that the direct visits had a higher number of last click conversions than assisted conversions. Display was proportionally the most assisting channel, although its number of conversions and value of transactions was small in absolute terms. It means that display banners on site and search engines brought a large amount of visitors who eventually converted. We can conclude that they are more efficient means of customer acquisition than other channels.
From this report we can also note that, while direct traffic delivered $171,259.73 of revenue based on last click, it also assisted in capturing $100,761.56 of revenue.
But this is only half of the story. To make real conclusions, we need Multi-Channel Funnel reports.
Top Conversion Path
The Top Conversion Path report shows us the path the visitor took prior to conversion. From this report we can see how many times, and through which channels, the visitor came to the site prior to conversion.
From here we can see that most conversions happened after two direct visits to the website. Most likely, the visitor came once, browsed through the products or just looked at the website, and then returned later to actually purchase something.
The path next in importance is Referral followed by Direct Visit. Most likely, the visitor first landed at the store through referral and purchased something on the second visit.
This list lets us segment our reports according to certain parameters to see how buyers behave and which behaviors correlate with shopping behavior.
Time Lag Report
This report shows us how much time passes between the first user interaction with the website and their conversion. Since this is an e-commerce store, we’re going to take it for granted that the sole purpose of the visit is to make a purchase, which explains why the majority of conversions happen on the first day of the purchase process.
But not all e-commerce stores work this way. Some require a longer funnel – a longer time between the first interaction and final purchase. Big-ticket items like mattresses, for example, generally involve consumer research before making the commitment. You’ll have to look at your own metrics to determine what an average funnel length is for your customers.
Path Length Report
The Path Length Report indicates the number of times the visitor ‘touched’ the site prior to the conversion.
This report indicates that most conversions are achieved by the first interaction with the site, and that virtually every conversion was achieved by the time the visitor hit the third interaction. This indicates that business cycle and conversion cycle are relatively short and that most visitors will either buy in their first visit or not buy at all. Having this information, we may safely conclude that remarketing efforts should be concentrated to the visitors at their second or third interaction, which would increase the proportion of those that eventually buy. Further down the timeline, likelihood of conversion decreases.
It also helps us determine the length of our time decay in the time decay attribution model. Having this information, we may conclude that time decay should be shortened from seven days in default model to three or four days.
All of these reports enable us to appreciate the contributions of the individual channels and assign them some value.
Note: We don’t have to depend on Google Analytics’ default last click attribution model. We can develop our own models based upon the insights the multi-channel funnel report yields.
One of the foremost experts (and practically the spiritual father of web analytics) Avinash Kaushik recommends that, in developing our own model, we should start with the time-decay model.
In the time-decay attribution model, the touch points closest in time to the sale/conversion get most of the credit.
In his (and our) opinion, this model is a bit more accurate.
However, it is not possible to use it without some modifications. Check out the series of posts on this on Avinash’s blog before setting out to develop your own attribution model.
Setting this up is not easy and requires a long-term commitment. However, once you manage this, you will be able to waste less money on ineffective marketing channels and use the most profitable channels to bring the visitors to your site who are more likely to convert. The hard work pays off!
What We Still Don’t Know
Multi-channel reports give you a lot of information, but not necessarily all the information. You’ll always have to take into account that some visitors will hear about your website offline. There’s no way to track that, unless you use QR codes or user surveys, but even then you won’t get a 100% accurate picture.
We also can’t be sure if the visitor linked all of the devices they use to a single account, which means visits may be recorded as coming from multiple individuals, when in fact it may only be one person, using multiple devices.
There is little to do to prevent this, but we can still gain incredibly useful insights from the channels we can track.
Google Analytics Multi-Channel Funnel Conclusion
Multi Channel Funnel grouping is essential. This is the most important of the above reports and you can use it to identify the channels that are not receiving their fair share of credit for conversions. Using these reports, you can adjust your channels and use your marketing budget wisely and efficiently.
When interpreting your data, keep in mind:
- The Visualizer tool is mostly just for show. Don’t stare at it for hours. Just consult it as a sanity check.
- You can spend as much time as you want on the Top Conversion Path, but unless your website is tightly structured with a limited number of entrances, you will have too many potential paths the visitor can take from there, which makes this report virtually useless.
- The Path Length report and Time Lag reports are very important for attribution modelling. Use the data from here to model your time-decay and position based attribution credits.
Google Analytics Multi-Channel Funnel Checklist
- Ensure your goals are properly set up. Without it, the multi-channel funnel report will yield no data.
- All the promo emails emanating from the website to customers or potential customers must contain links that are UTM tagged. UTM tagging means adding attributes to the links that tell Google Analytics more details on the channel and medium the traffic came from. UTM tags enable Google Analytics to report the name of the promotion campaign so you can attribute conversion to your marketing efforts. There is a simple tool – Google UTM builder – that enables you to create your own custom UTM tags.
- All the links that are part of promotions/ads on third-party sites (like Facebook, Bing, etc.) should also contain UTMs.
- Ensure that all the channels are grouped by using your view settings. This way, you can name each of your channels and mediums in a way that gives you more information than Google’s defaults.
- Use common sense when looking at your data. Cross reference different reports to spot possible anomalies. For example, check landing pages, compare number of visits with and without transactions using custom segments, and so on. Be sure you are not filtering any channel by mistake.
This is one of the important reports in your Google Analytics account. Be sure to use it. It may save you some money or enable you to allocate your marketing budget in a way that’s more likely to bring visitors with higher conversion potential to your website.