It’s time for part two of our three-part series on getting the most out of your Salesforce Marketing Cloud Intelligence, or Datorama, instance. Part 1 of the series touched on little tricks and shortcuts to success with data ingestion and transformation. Today, I’ll be going through how you can speed up the platform for yourself and your team and boost efficiency in reviewing your data. Read on for Marketing Cloud Intelligence tips that will help you maximize its performance.
Marketing Cloud Intelligence Performance Boosts
Marketers are competitive by nature. You’re fighting for bids, placements, and user attention. So why wouldn’t you want the best performance out of the tool you’re using to measure your data and give you insights?
We’re going to walk you through how to ensure you get the speediest and most efficient version of Marketing Cloud Intelligence you can!
Less widgets, more pages – why?
You may be inclined to make a gigantic page of widgets for one humongous dashboard that you use — this is a fine starting point! It’s especially useful for the sake of exporting data in a PDF or PowerPoint in an advanced manner, using the platform’s advanced export feature.
But when it comes to using the actual dashboards in Intelligence, you should try to break your pages out as much as possible. This could be done by channel, campaign, or any other kind of unifying but specific factors you might be thinking about.
We’ll touch on this more below, but the more compact your view and the fewer widgets you have, the less you’ll wait for pages to load.
A load by any other name
You may have sat in front of numerous online analytics tools and seen long loading times. You probably used the old answer as to why: slow internet speeds.
Source: https://i.imgur.com/us5fOfG.jpeg
The true answer to why your analytics tool is slow to load though is… it depends:
JavaScript ifs VS. spreadsheet IF formulas
I’ll cite a specific example below, but any calculated fields may slow your load down. I noted some of this in my first blog in this series, but using JavaScript ifs can help a bit by setting quick end criteria (ie. if it’s this thing, this is the output), as opposed to large loops in Excel style IF formulas.
An additional point of clarity: when I note calculated fields, I mean calculated measurements, filtered measurements, calculated dimensions, and patterns.
Data fusions & transitive fusions
Data fusions similarly happen live and can significantly slow down your instance, either with heavy data sets or more than 5-10 fusions existing back end. Transitive fusions (a is fused to b, b is fused to c) can also cause problems in heavy use.
Page reloads
Accessing a page in a way that makes data load anew can cause slow load times.
- I’ll hit some rules of this later, but scrolling down will cause data to load anew
- Any updates to the data back-end will cause reloads of the data when you open a page with that data
- Clearing your cache either in the platform or in your browser will cause data to recalculate and populate the dashboard page
- Changing filters will also reset the loading of the page
Identifying load problems & getting help
There is not currently a public-facing timer for loading data in Intelligence, but you know a long load when you see it.
What can help in identifying the root causes of slow loads is that Salesforce support can help diagnose timing issues down to certain widgets. I recommend this (or the next method) to validate that your internet/instance is not the issue. You could also have a colleague access the same dashboard and compare load times with you using a timer to ensure you are capturing the issue correctly.
Come Together (but also maybe don’t)
One of my favorite functions of the Intelligence platform is using Coalesces and other calculated dimensions whenever I can’t cleanly join data. There are other ways to do this — like using if statements with ISEMPTY formulas, using data fusions, or doing Vlookups to tie entities together, among others (we’ll touch on some of these below, too).
Coalesce is often one of the cleanest and quickest ways to join data points where you need a data point from one stream or another, depending on what’s available. A good example is something like lead source, where you may have leads and cost data to join across Salesforce data and paid media APIs. In that case, you would check the lead source field or the channel name of the data stream feeding cost data).
Technical tip sidenote: Coalesces can be incredibly confounding, so here’s a brief walkthrough of this function. In this platform, as documented in Intelligence’s Help Center, they let you return the first non-empty value of a set of data formatted like COALESCE([value1, value2,…]). The really confusing part is that you have to have brackets around those dimensions, so correct syntax would be like COALESCE([[day],[campaign start date]]).
Scroll first, ask questions in a minute
One thing that can be really frustrating, even when things load okay as a page opens, is that as you scroll the content may not be there when you move down. A way to avoid this is to scroll all the way through a page before walking through it.
Why is this? Two big reasons:
- Intelligence loads about 15 widgets at a time. Anything more than that queues up, so multiple widgets loading piece by piece can be difficult on the system.
- The system only loads what you see. There is no secret back-end loading happening wherein data is populating that you don’t see. You’ll still have to reload widgets when you filter, but by scrolling down, any initial data sets will populate and make you appear to be the marketing superstar we know you are.
Filter secrets-greater than 0
- Have you ever gone into a page and waited for a filter to load?
- Or, have you opened a filter widget’s dropdown to find hundreds of irrelevant campaigns when you know your workspace should only have a handful of campaigns?
I use a trick that I call “Greater than 0,” which is a calculated measurement that adds all of my measurements in use for a page and filters that on greater than 0 for any relevant filters. My filters load faster and my dropdowns are minimized to just data points I need, not junk data.
You’ve successfully loaded the end of this article
Thanks for scrolling down here successfully and ingesting our content! Now you have quick wins that can help you maximize Marketing Cloud Intelligence performance.
Check out part one of this Marketing Cloud Intelligence series to get a reasonable foundation on some other data topics. And keep an eye out for part three of this series, which will go in-depth on some data stream tricks to get your best data pulls/updates possible.
We hope you get some added productivity out of this piece, and if you’re looking for a tailored conversation to your needs, drop us a line.
