- Improve keyword coverage to reach customers at all stages of the conversion path.
- Identify those channels that directly contribute to the growth of your business.
- Learn how metrics like average order value can be influenced by early-stage marketing.
What do I need to use Multi-Channel Funnels properly?
MCF Implementation Checklist:
- Install Google Analytics! Make sure that all of your webpages are tagged, and if you happen to have more than one website (yoursite1.com and yoursite.2com) or multiple domains (red.mysite1.com and blue.mysite1.com) that you are set up to use Multi-Domain tracking. This last step will ensure that you are tracking all interactions across your sites into a single customer path.
- Set up E-Commerce Tracking or Goals. MCF needs to know what action represents the very end of the customer path - the conversion. The conversion may be a sale, or it could be another action that’s valuable for your business, like filling out a lead form or downloading a brochure. For businesses selling products online, you can measure conversions (sales) through e-commerce tracking. If you’re measuring visitors that take a specific action, such as completing a form, setting up goals will suffice.
- Get your tags in order. For AdWords customers, make sure that your advertising account is linked to your Google Analytics profile and that auto-tagging is enabled. For other channels, such as e-mail or advertising run on other networks, our custom URL builder will help you build the tags necessary for each campaign. If you’re new, be sure to learn more about channels and channel grouping.
- Start using the MCF reports. Once you’ve followed the steps above, you can find the Multi-Channel Funnels reports in the Standard Reporting tab of Google Analytics: click on “Conversions” at the left-hand side of the user interface, then click “Multi-Channel Funnels.”
Is it possible to integrate the data from Multi-Channel Funnels directly into our own systems?
Absolutely. Not only are all of these data points available for export from the Google Analytics interface in commonly-used formats, we also just announced the release of the Multi-Channel Funnels API so that developers can tap directly into this incredibly powerful data source. See our recent blog post for more information.
How do we ensure we are tracking all our channels in a way that is optimal for these reports?
By default, all inbound clicks that are part of a conversion path are captured by Multi-Channel Funnels. The default channel groupings that we provide then make a series of fairly reasonable assumptions to group traffic into their respective buckets. As a user, you have two approaches to ensure that all traffic is ending up in the right place:
- The first option is tag all of your marketing activities in a way that matches the logic of the default channel groupings. You can find the rules behind the groups in this help center article. There is also a simple URL builder so that you can append the proper tags to your other campaigns.
- The second option is to create channel groupings that match the way you are currently naming and tagging your campaigns. This approach tends to be favored by those companies that want to utilize all of their historical data in MCF right away, or have technical limitations preventing them from changing the actual campaign tags. Implementation details for this approach can be found on the Analytics Help section, in this article on channel groupings.
Does MCF have to be a true purchase or it will it work for a Business-to-business company looking for qualified leads?
Companies that are pursuing leads tend to have much shorter conversion paths than those that are tracking purchases. It's not entirely uncommon to see lower assist / last ratios and, equally, to have the perception of less opportunity when reviewing the MCF reports around a single goal. As a better practice, we suggest that advertisers implement multiple goals to measure customer activities along a wider path.
For instance, goals could be set up at points before filling out a lead form but after becoming a slightly more qualified customer, either by increasing time or page depth on your website, reviewing a whitepaper or looking at cost information. These would help to measure performance even if there is a more significant lag before becoming a lead, lending insight to the very early parts of the journey.
After the lead form is filled out, any unique action that you could encourage to bring the now qualified customer back to the site again, such as completing a signup process, reviewing a contract or qualifying for a promotional offer, can then be used to go all the way back through to the beginning of the journey to find that initial contact point.
Why is (not set) so high for AdWords Keyword?
When you select a primary dimension in the Assisted Conversions report of Multi-Channel Funnels, it is not filtering the information as much as it is adding a different view to it. As such, when I move from that basic channel grouping view to AdWords Keyword, the report still shows 100% of the data but now groups each interaction by its respective keyword. However, since not all interactions have AdWords Keyword data associated with them, including Direct, E-Mail and Social Network visits, they are grouped into their own (not set) bucket.
During the webinar, my colleague responded to this question by saying that “not set” may also appear due to broken AdWords tags. This response is also technically correct as broken AdWords tags can also prevent keyword information from being passed through, but in many cases it’s more likely that it’s just because the visits don’t have keyword data associated as described above, and AdWords tags are probably OK -- so consider this first before trying to troubleshoot.
What devices are in place to prevent Spiders and Bots from inflating data and thus causing a possible "bad" business decision?
Multi-Channel Funnels measure specific goal or conversion actions that are hopefully beyond the grasp of bots or spiders that are just mining content. For instance, it probably wouldn't be likely to find one that tries to fake e-commerce orders.
If you have found bots coming through these conversions on your website (i.e. Store Locator), it may be practical to filter those visits out at the profile level in Google Analytics to make sure that they are not impacting any of your resulting analyses. Although we don’t recommend a specific set of criteria for limiting bots, there are dozens of articles online that you should be able to find with individual opinions on what is best.
Posted by by Neil Hoyne (@nhoyne), Global Program Manager for Attribution