Caucasian advisor financial business analytics woman with data dashboard graphs

The Trusted Advisor Roadmap Part 2: Building Attribution Models

Once you’ve helped your client define their event goals, the next step is to determine a means of measurement, which can be an even more difficult process. Here, again, an imperfect measurement metric is better than none, so let’s explore several ways to measure success.

In the last article, I used the example of the Industrials Group at an investment bank wanting to host a conference, with four goals identified. The goals were weighted to assign relative importance, and the top one was “close deals with existing prospects.” Let’s ignore the other goals for now, and focus only on that one.

Let’s assume the event costs $400,000, and that an average deal generates a fee to the firm of $2 million. The next question might be, how many deals do we need to close? A quick calculation shows that even if only one deal is closed as a result of the event, the conference will have paid for itself five times over, while closing, say, eight deals would appear to be a home run, right? Well, not so fast. It’s highly doubtful that any deal will close solely because of the event. Bankers probably spend months, if not years, courting their clients. Let’s agree, then, that we can attribute some percentage of the deal closing to the conference, but not all. The question is how much. To determine this, we need to build an attribution model.

1.  What Is An Attribution Model?

Attribution models are used to assign a percentage of the credit for a sale to the various stages or touchpoints a customer goes through to get there. They’re most often used in digital marketing for new customers who may have multiple touchpoints with a brand (search results, paid search ads, social media posts and ads, website clicks, etc.) and Google Analytics offers several types of attribution models to choose from (e.g. Last Interaction, First Interaction, Linear, Position-Based, etc.)It’s much harder to build attribution models for high-service, long-sales-cycle, B-to-B transactions like investment banking, however. But we can try to attribute a percentage of the sale credit to the conference, however imperfect it may be. Only your client can agree on what that assigned percentage is, but you can help them by presenting several scenarios.

2. Sample Attribution Model for an External Event: Investment Banking Conference

In the grid above, I’ve run calculations showing the event contributing anywhere from 1% to 40% of the client’s signing, as an example. This is far from an exact science, of course, but you have to start somewhere, and coming in with a range helps get the conversation started. 

You’ll want to pick numbers at each end of the range that you know are somewhat unreasonable, to get your client to establish realistic parameters. If in this scenario, for example, your client says, “Come on, we know the conference contributed more than 1% toward closing these deals, but 40% seems really high.” Now that you’ve got buy-in that the attribution percentage for the event is somewhere between 1%-40%, you can ask what would be revised parameters to re-run the projections with.

In Column A, I’ve also run the scenario for different numbers of deals closed, from two to twelve. The calculations in columns E-I are arrived at by multiplying the Total Revenue X the Attribution %. Presenting a matrix like this, with variations of both the numbers of deals closed and different attribution percentages, can help you and your client find a “sweet spot” of success for the event.

For clarity, I’ve also used the Conditional Formatting feature to have the cells automatically highlight in green if the calculated attribution figure is greater than the event cost, and red if it’s less than the event cost.

This doesn’t mean if the cell is green that the event has a good ROI. As we’ll tackle in another post, ROI is a relative concept.

Be prepared to also have a conversation about defining event costs. Planners typically focus on external expenses like the venue, catering, audiovisual, etc. However, the internal time of the planning team needs to be taken into account too, as well as the time the bankers will spend preparing for, and attending, the event. You don’t have to necessarily include these costs in your calculations, but it’s important to show your client that you’re aware of them.

3. Sample Attribution Model for an Internal Event: Sales Incentive

Now let’s try an attribution model for an internal event such as a high-end incentive trip for top salespeople. Let’s say an organization has 100 salespeople who average $400k in sales per year. The head of sales has a budget of $300k to spend on an incentive trip for top performers, to motivate the entire team to increase sales. In the matrix below, Column C lists several different percentages of increased sales, while Columns G-I model out scenarios of how much the incentive trip contributed to that increase. There are numerous factors that affect sales, including the economy, competitors, changes to the industry, etc., and the purpose of the attribution model is to identify how much of an impact the event has. 

If sales increased by $8 million, or 20%, for example, and you agree that the incentive had a 10% impact on that, the calculation shows $800k of increased revenue attributed to the incentive (cell H8). Here, again, the parameters will need to be hammered out between you and your client, but this provides a jumping-off point.

4. Conclusion

Attribution models are useful in assigning how much value an event had in contributing toward a specific goal. They are never an exact science, but, as with most event ROI, an imperfect model is better than no model at all. More importantly, attribution models can be a vehicle to engage in a strategic conversation with clients about how the events you work on drive broader business goals of the organization. And they will come in handy when we get into calculating ROI in future articles in this series.

To go deeper into this topic: