The user interface required you to know to click a specific button in a specific part of the window and there was nothing in the workflow or intuitive about the user interface that would lead you to take that action. The terminology on the screens is unique to Marketo and has little relevance to Salesforce.com terminology. The more I deal with Marketo, the more I feel that the user interface is almost as poor as the user interface in Eloqua. I would give Eloqua an F for its user interface and Marketo a D for its user interface. Now I understand why Marketo wouldn't let us do a free trial of their product because we would have realized that their "ease of use" is not that good.
Back to the Analytics. If you want to do some basic reporting, Marketo Analytics and Subscriptions are fine. Subscriptions is simply Marketo's term for a report that runs on a regular basis and is automatically emailed to you. In general, most of the reporting that is available in Marketo can be replicated and with better detail in Salesforce.com. What Marketo is good at is looking at individual lead, contact, or opportunity behavior. However, if you want to do any analysis of groups of leads, contacts, or opportunities to identify trends and averages, then Marketo doesn't do that very well.
We tried to look at two analyses and Marketo's reporting on both of them were not very useful.
1) We wanted to run a report on new leads that came in from the website, what date they were created, and what specific content or form on the website generated the lead. Marketo could easily show us how many website leads were created and what was the source of each lead. However, when we tried to get the system to show us the create date in addition to that data, Marketo couldn't do it. In salesforce.com, I would simply run a report with Create Date, Lead Source = Website, and Lead Source Name to get that information. Marketo was unable to break down the data along those dimensions. It could only generate a table showing the aggregate leads in that timeframe by the source of each website lead.
2) This is a more complex analysis, but my expectations had been set that Marketo could export this data into a spreadsheet where I could analyze it. I wanted to do an analysis based on engagement level. This engagement activity analysis can be used to segment your database to identify leads that are highly engaged and leads that have low engagement so you can potentially drive different marketing programs based on a lead's level of enagement. Example, for someone that is unengaged, then send really interesting subject lines just to try to get them to open an email and click through or send them a survey to try to understand what they're interested in.
So we tried to run a report to identify all leads based on number of email clicks that they've done. Example: Identify all people that have never clicked thru on an email, all people that have clicked once, all people that have clicked twice, etc. The only filter that was available in Marketo do this analysis was "clicked email a minimum of X times." So if I want to get a list of everyone that has clicked only once, I would have to run three reports (everyone clicked mininum zero times, everyone clicked minimum 1 time, everyone clicked mininum 2 times). Then I would have to use Excel VLOOKUP to compare all three spreadsheets to find only the people that have clicked one time. Our database is so big that doing that analysis with 200,000 records would be painful.
Here are a couple other issues with Marketo reporting/analytics that people should be aware of:
- Marketo doesn't have a way to filter out data based on IP address. What that means is that when you're visiting your landing pages and clicking emails to for testing, your visits and clicks are going into the Marketo reporting. Marketo doesn't have a way to filter visits from your IP addresses from their database. What that means is that your visit data is going to be inaccurate. Marketo's position was that your number of visits will be small relative to your total visits so the level of error should be acceptable. Really disappointing that you can't filter out your testing from the Marketo data.
- It's not practical to analyze the data in the Marketo Activity Log. Marketo has an activity log that tracks every activity on a lead record including clicking emails, site visits, downloads, and field value changes. So if an address is updated or information is entered into a field, that goes into the activity log. This is great if you want to look at an individual. The activity log is useless if you want to do any analysis to identify trends. Example: Analyze all Leads that converted into Contacts over the last 30 days to identify what activities drove conversion across those people. Marketo is designed to look at each person individually. It doesn't have the ability to analyze that group of people to identify any trends. So I run the Activity Log report and it generates 100 people that have converted from lead to contact over the last 30 days. It provides a list of each person. You can click on each person and will display the detailed activity log for that person. The activity log contains every single change that occurred to that lead. So you have field value changes in addition to activity. There's no way to export that list for all 100 people in the report. What I would have to do to do any trend analysis would be to export the activity log as a spreadsheet for each individual resulting in 100 separate spreadsheets. Then I would have to combine all 100 spreadsheets to be able to do the trend analysis that I'm interested in. You could do it, but it's not something that you would want to do ofter or with large numbers. I was expecting to be able to export the data for all 100 people into one spreadsheet and to be able filter out field value changes - not possible in Marketo.