In our prior post on Measuring Social Media Impact we briefly mentioned Multi Channel Funnels and promised to discuss them soon. This post will discuss Multi Channel Funnels and what they can tell innkeepers about reservation sources.

What are Multi Channel Funnels?

[pullquote]A significant number of visitors make multiple visits before booking. That, alone, should make us pause before we draw conclusions about the value of a referral source only from the number of visitors it sends.[/pullquote]If you’ve used Google Analytics (GA) for enough time to be comfortable with it, you’ve either realized or learned from others that one problem with much of the data it relies on is that it is “last click” data – GA identifies the source that sent the visitor to your website, but historically didn’t credit any other source that may have sent the same visitor at an earlier time.

Nevertheless, we in the B&B industry know that nearly all visitors make multiple visits to your website (as well as other sites) before booking. Google knows that, too, and, in August, 2011, released Multi Channel Funnels (sometimes known as MCF) to help identify the sources that contribute to a goal conversion. In the B&B industry, a product has existed to record these visits for quite some time: Intell-A-Keeper, from Acorn IS provides similar information about multiple sources contributing to a booking.

When discussing Multi-Channel Funnels, we speak of conversions (that is, a visitor having completed a goal you have set up), rather than bookings, because Multi-Channel Funnels are based on Goals (and you must have at least one goal set up in Google Analytics to use Multi-Channel Funnels). In addition to having a Goal for an online booking, Multi-Channel Funnels can also use Ecommerce Transactions as Goals.

What Can I Learn from MCF Reports?

There are five new reports that use MCF. To find them, use the left side navigation to go to Conversions -> Multi-Channel Funnels, where you’ll find:

  1. Overview
  2. Assisted Conversions
  3. Top Conversion Paths
  4. Time Lag
  5. Path Length

The Overview page shows Total Conversions (including conversions that would not qualify as MCF conversions, because GA sees only one referral leading to the conversion), Assisted Conversions (where multiple sources produced visits leading to a conversion), and a “Conversion Visualizer” – a Venn diagram of overlapping circles, showing the separate conversion rates of different channels, and the overlap showing where there are assisted conversions.

The Assisted Conversions report groups channels in familiar GA groups – Organic search, Referral, Direct, etc.

The Top Conversion Paths report shows the sequence of channels leading to a conversion (from Organic to Direct, for example).

The Time Lag report shows the length of time from the first visit to the conversion (currently limited to 30 days – if the time span is longer, the report will not include earlier visits).

The Path Length report shows the number of visits from different channels leading to a conversion.

Putting MCF to Good Use

Setting date rangeFor smaller properties, as is often the case with GA reporting, there may not be enough traffic in the default time frame of one month to get meaningful information. The simple way to increase the amount of information is to select a longer time, by going to the top of your analytics and using the pull-down arrow to select a longer date range. You could try a year, or several months, for example.

One of the most useful reports for innkeepers will be the Assisted Conversions report. After selecting that report (Conversions -> Multi-Channel Funnels -> Assisted Conversions), click on the colored Referral line under Basic Channel Grouping. By default the resulting list is sorted by the number of assisted conversions participated in by each referral source. There are two other very interesting pieces of information here.

Recall that the number of Assisted Conversions for a referral source is the number of conversions this source participated in (where there were other referral sources also). The first additional piece of information is the number of Last Interaction Conversions. These are conversions where this source was the last referral source before the conversion. GA seems to consider these “closers” – sources that close the transaction.

Assisted/Last Interaction RatioThe second piece of information is the last column on the right – the Assisted/Last Interaction Conversions. This column represents the ratio of Assisted Conversions to Last Interaction Conversions. This provides some interesting insights. If this ratio is less than 1, the referral source is a “closer” – one that is more often the Last Interaction source. If the ratio is greater than 1, the source is an awareness generator. A ratio of 1 indicates a source that plays both roles.

Armed with this information, we are in a much better position to consider whether a particular source (let’s say a B&B directory) is sending us visitors who book, than by simply using the number of visitors from that source in the All Traffic or Referrals reports, or by relying only on the Ecommerce or Goals data. We may find that a directory we thought wasn’t producing bookings is in fact acting as an awareness generator, even though other sources send the Last Interaction.

Path LengthIn a similar way, we can examine the Path Length report to see that, while (in this example) 43.7% of visitors convert (book) in one visit (but see below for qualification of this), but that means that 56.3% are coming from multiple sources on multiple visits. That, alone, should make us pause before we draw conclusions about the value of a referral source only from the number of visitors it sends.

Creating Custom Channel Groupings

The MCF reports also allow you to create your own custom channel groupings. We’ll post about that in the future. However, if you’re not happy with the default groupings, there are tutorials to create your own here and here.

Some Words of Caution

Because the reports we get from GA (or any other source) are only as good as the data they rely upon, they are necessarily imperfect. For some kinds of information, such as the reports discussed in this post, MCF reports are a great improvement over straight Referral or Goal Conversion (or even Ecommerce) reports. However, they remain imperfect.

For one thing, as mentioned above, the data is currently only considered if it is within the past 30 days. If a visitor books, but their first visit was 31 days (or more) earlier, that first visit (or visits) will not be included. For another, many people use multiple devices to search, so the cookies that GA relies on actually reside on more than one device, and GA has no way to identify which belong to the same visitor and aggregate them. To be fair, we know of no stats package that can successfully track a visitor who is using multiple devices, without requiring that the visitor sign in every time they visit the website.

Imperfect though the information from the MCF reports may be, in many ways it is a great improvement over simple referral and conversion information. It certainly adds more and valuable information to the decision to use, retain, or quit a particular referral source.