Why You Need to Focus On Defining Campaign Attribution in Salesforce

One of the most importing aspects of marketing is knowing how your campaigns contribute to revenue. This knowledge allows you ascertain which marketing channels or campaigns are most effective and helps you allocate future spend and resources effectively.

Today, we have access to an untold amount of performance data — clicks, displays, email opens, online conversions, downloads, time spent on a webinar, content downloaded, people scanned at an event — I could go on all day. So its never been easier to determine ROI for your campaigns, right?

The truth is Campaign Attribution is hard. There are many different models, including more advanced ‘Multi-Touch Attribution’ which weights every marketing touch on the way to a sale — or more basic models like Last Touch or First Touch which are significantly easier to implement and manage.

There is no one-size-fits-all for campaign attribution models and the route you take, will be governed by the type of business and campaigns you run.

Here I talk about the models I use day-to-day and how they help me make better decisions. Later in the article, I highlight some drawbacks with my model, which should help you consider the best options for defining campaign attribution for your business.

The Campaign Attributions I track are:

- First Touch (or Lead Source)
- Tipping Point (or Primary Campaign)
- Campaign Influence (arguably Even Split)

First Touch

I use First Touch to determine the source of all Prospects entering our funnel. This metric tells me where to spend money to gain net new leads. The First Touch applies to 100% of Lead and Opportunities, regardless of whether they are Won, Lost, Working or Unqualified. Everyone came from somewhere! In Pardot, Campaigns are a prospects first touch. To ensure you measure this correctly, you can follow my guide on How to use Google UTM parameters with Pardot.

Tipping Point

When it comes to B2B or Account Based Selling, you inevitably run lots of different campaigns during a prospect’s lifecycle. I, however, find that there is normally a specific event that occurs which launches the prospect into ‘sales ready’ mode. This could be an event, a specific email campaign or even a promotional offer. The Tipping Point tells me what campaigns directly contribute to pipeline contribution. The Tipping Point applies to around half of the opportunities in our company pipeline. The rest is made up of inside sales, outbound sales and partner referrals.

Campaign Influence

It’s true that we use Salesforce’s Campaign Influence Report to work out the Tipping Point (or Primary Campaign) of closed won opportunities. However, I also use it to look at the overall return of our marketing campaigns.

My Campaign Attribution Models in Practice

Let me now give you a real life example of how these three models work together (using completely imaginary data of course!).

Dreamforce Sponsorship 2016
Each year, we sponsor Dreamforce and run a range of campaign activities. For the purpose of my return on investment reporting, I set up two Primary Campaigns: Dreamforce Expo ’16 and Dreamforce Party ‘16.

Event Sponsorship: £20,000
Swag, promotions, giveaways: £10,000
Flights, hotels, subsistence £10,000
Networking Party £10,000
Microsite, newsletters £2,000
Paid Advertising £3,000
Total Spend £55,000

Marketing Conversions
Pre-Event Networking / Party Signups
800 people / 250 new prospects
Pre-Event Meetings Booked 80 people / 0 new prospects
People Scanned on Expo Floor 1800 people / 960 new prospects
People Scanned at Parties/Networking 1200 people / 600 new prospects
Total 2600 people / 1500 net new prospects

Note, the total number of people and leads do not add up because people are in multiple categories.

3 months later I’m considering whether or not to sponsor Dreamforce 2017, so I look at my Campaigns in Salesforce.

So a few points to mention here. I’ve split the flight costs etc evenly between the two events. I have also prioritised the Dreamforce Expo for Opportunity association over the Dreamforce Party (e.g. if prospect was at both, the Expo is attributed).

Primary Campaign: Dreamforce Expo ‘16
Cost: £40,000
Campaign members: 1800
No. Opportunities: 400
Pipeline Contribution: 200,000
Total Value Won: £120,000

Primary Campaign: Dreamforce Party ‘16
Cost: £15,000
Campaign members: 2000
No. Opportunities: 10
Pipeline Contribution: £50,000
Total Value Won: £1000

From the above it’s pretty clear that we should do the Expo and stop wasting money on partying!

But now let’s look at Campaign Influence

Campaign: Dreamforce Party
Campaign members:
Number of paying customers: 400
Total value of customer ACV: £120,000

By the looking at Campaign Influence we can see that whilst our Dreamforce Party did not offer anywhere near a full return on investment, it gave us an amazing opportunity to network with customers. Therefore, it could be argued that this was a worthwhile expenditure and should be repeated.

Fast forward 3 months…

January was a record month. We hosted an event in San Diego that brought in 5 new deals and £60,000 of new revenue (in this case, the event is now the Tipping Point / Primary Campaign).

Where did the prospects originate from for this new revenue?

By analysing the First Touch of the pipeline generated by this new Primary Campaign, we can see that our Dreamforce Party is showing greater returns than initially anticipated.

I have tried to keep this simple for the purpose of getting my model across. Hopefully you can see that by analysing First Touch, Tipping Point and Campaign Influence together, I can ascertain the broad extent of ROI from each campaign.

Campaign Influence Drawbacks

It’s worth mentioning that life isn’t quite this simple. What happens if you do a multi-channel campaign and a single prospect comes to an event, clicks on a paid Twitter advert and is also provided as a paid lead from somewhere like G2Crowd or Capterra? Who do you attribute the lead to?

In addition, Campaign Influence doesn’t scale down or provide deeper insights into the meta data of your marketing performance - this ultimately leads to siloed reporting processes.

You may have noticed the line item of “paid advertising’ in my Dreamforce budget. In terms of Salesforce campaign reporting, this cost was simply lumped into the overall campaign. Now, you may think I could have divided out those people generated from my paid campaigns and set up a separate Salesforce campaign…

The question is how far do you go?

Am I going to set up a Salesforce Campaign for prospects that were gained from: all paid? paid versus channel? paid versus source? paid versus advert content? paid versus medium? (display, cpc cpm etc)

The truth is, in order to accurately measure marketing, you have to go beyond capturing the interaction with gated or uploaded content and look at meta data. Whilst we can use meta data to measure the Source and Content that generates a new prospect and view that in Pardot, it’s impossible to gain real insight to the full journey that prospect takes on it’s way to a sale in Salesforce.

For example:

We can understand:
- What content generates a first visit?
- What content generates a conversion?
- What events, webinars and campaigns do prospects engage with?

We can not understand in a Salesforce Campaign:
What is the optimum product video minutes watched to gain a sale?
Which content formats work best at last stage of the funnel?
What content within paid campaigns produces the most pipeline?
Which content writer generates the most readership from C-Level execs?

Consider what you need and will use

For me, you really have to weigh up how much marketing spend you are allocating versus the return you will get from implementing fully fledged content marketing analytics. If you are producing less then 20 content pieces per month and you have a good hold on Google Analytics reporting, you can get by without diluting marketing spend with ‘operational’ costs.

If you can get to a level where you are reporting on everything I have discussed in this post, then you will be in a great place to make an educated decision on whether that reporting is sufficient for your needs.

That said, if you run lots of multi-channel and multi-device campaigns or invest heavily in content and paid online marketing, you should probably be looking at platforms that support Multi-Touch Attribution with Meta Data.

To conclude

Marketing needs to measure and prove ROI. The campaign attribution model you employ, will dictate how that ROI is measured. If you haven’t already, you need to focus on defining campaign attribution for your business, before running campaigns you can’t consistently report on.

That’s it. I’d genuinely love to hear your own views…

Other resources that you may find useful:

An Introduction to Multi-Touch Attribution by Segment

Multi-Channel Attribution Modeling: The Good, Bad and Ugly Models