Thursday, December 25, 2008

Back to Basics: Calculating ROI

Most of the marketing positions I've held have required me to demonstrate how my department's activities have produced a significant return on investment (ROI). I've learned that this seemingly simple concept is not as cut and dry as it might appear - even if we leave social media marketing and perhaps even traditional advertising out of the equation. Yet demonstrating the value of marketing activities has always been how I've proved that what I am doing is worthwhile.

Jim Novo, author of Drilling Down, makes the point that because there are so many ways of calculating ROI, the best way to calculate ROI is to use an equation that your company (ie your Board or CEO or CFO) finds relevant. How true. (You can read Novo's full article here.) Perhaps this should be step one in any quest for demonstrating accountability in the marketing department - determine what metrics the big bosses want to see.

In the earlier parts of my career, I did just that. When I calculated ROI, I took into account only the cost of acquisition and the resulting profit of a campaign (this was all that my "big bosses" wanted to see from me for a long time). As I progressed and became the manager of a department, however, I had to account for additional factors including the cost of staff salaries (my "big bosses" became even bigger bosses - CEOs and the Board and the like). My understanding of ROI changed as a result, and has continued to change with increasing responsibility and "higher" perspectives. The "higher up" I go, and the more I reflect, the more and more factors I believe I need to take into account. The more I learn about measurement, the more I believe that the results of the calculations I have relied upon only give me a glipse into the effectiveness of my marketing activities, and therefore how limited I have been in my ability to substantiate the value of marketing activities that relate to longer-term profitability goals.


The type of ROI calculation I am most familiar with - ie profit/cost of acquisition, is what Novo refers to as "simple ROI". As Novo points out, although this type of calculation is important, the results demonstrate only "front-end profit." These results ignore other important factors like the effect on the brand, customer loyalty and repeat purchases. According to Novo, these are all factors that can be measured, too, but the calculations are somewhat more complex. I'll delve into these areas in due course in this blog. I have always known that these factors were missing, but to date I have been unable to really find adequate methods for quantifying marketing's contribution to these areas.

In regards to web metrics, Novo suggests a very interesting model. It seems rather straightforward. In order to calculate the ROI of push campaigns where information is gathered through a company's website, especially where the campaigns are complex, Novo suggests that we might examine the revenue over any given period divided by the number of, let's say, unique visitors to the company website to determine how much each unique visitor is "worth". For example, if my revenues over one week are $1000, and I have had 100 unique visitors to my site, then by this calculation, each unique visitor is worth $100 to me.

I think this is a useful calculation, but I'm not sure I'm satisfied with assuming that my revenue is directly related to the number of uniques to my site. I think there are too many factors at play for this to be an accurate measurement. Still, I think that demonstrating the relationship between increases in uniques and increases in revenue are definitely worth looking into. If these were plotted in a graph over several months, or preferably, over several years, this would yield a benchmark for the ideal number of unique visitors to help stimulate revenue. I find this intriguing. I think this might be the case for a number of different methods of measurement - they are incomplete on their own, but taken with other relevant metrics they become more useful.

In fact, the more I come to understand about measuring the effectiveness of marketing activity, the more I believe that any marketer worth her salt needs to create a system of interconnected metrics to attempt to show progress towards company goals. Granted, this picture will always be incomplete because we cannot measure absolutely every effect for want of time, however if we choose a variety of the most important metrics, we will surely arrive at a more complete picture... and that complete picture may just save our marketing budgets (or our jobs), and I am convinced will absolutely allow us to speak from a stronger position when the big bosses want to see measurable results.

Friday, December 19, 2008

The Quest Continues: Finding Ways to Measure the Effectiveness of Social Media

Once again in my search for concrete metrics for the measurement of participation in the social media sphere, I've come across some interesting material that touches on the subject just enough to whet my appetite, but not really give me any foundation for building a strong business case for a social media campaign. Perhaps I've spent too much time with CFO's and bottom-line focused CEO's; I still believe that if you can't figure out how much money your efforts are generating, why would you participate at all?

My most recent nugget of golden info comes from an interview with Lawrence Swiader, Director of New Media at the National Campaign to Prevent Teen and Unplanned Pregnancy as published by The Buzz Bin. (I'm not up for debating the moral basis of the organization by the way -- we'll leave that for someone else's blog... I want to discuss the metrics and only the metrics.)

Swiader notes that the success of their social media campaigns (they use all the usual culprits like Facebook, Twitter, blogs and the like), will be measured firstly by the number of eyeballs and ears they reach. Secondly, a "higher level" evaluation will be undertaken to somehow measure how much information has been imparted and then stored in some way (I assume this means via an online vehicle and not 'stored in the brain'). Swiader doesn't give any details as to what this second evaluation might entail, but perhaps that is because it hasn't been attempted yet. It still sounds promising.

What I like about this interview is that although no hard numbers or formulas are given, it has reminded me that campaigns, social or otherwise, shouldn't be about just "throwing it out there" and then seeing what the results are, but that there has to be an ultimate goal, and then progress may be measured in relation to that goal. For the National Campaign, that goal is to reduce unplanned and teen pregnancy in the US, and therefore their social media goal is to reach as many people in their target audience as possible. We may not all work in the not-for-profit sector (I don't), but this simple methodology should still apply.

So let's apply these same simple principles to a business model I am very familiar with -- the business conference/tradeshow industry. Let's say that my ultimate goal is to generate as much profit as possible. To do this, I know that I need to attract as many attendees and big dollar sponsors as possible. My social media goal, then, is to reach as many of the eyeballs and ears of my two target groups as possible, which will yield some quantitative results. Then it seems to me that the next logical step would be to substantiate these findings with some sort of qualitative report that describes the quality and depth of the conversations and not just how many times something was read -- afterall, activity is an indication of interest. The combination of qualitative and quantitative results should provide a relatively accurate picture of the effect of my social media campaign.

But we can't stop there. Because in the conference world, and many other businesses, there remains an important question -- what does this mean in terms of bucks in the bank? For the National Campaign, it's important that they decrease the percentages of unplanned pregnancies. In some ways this is not dissimilar to ROI or revenue. At the end it's all about how the metrics you choose translate into a business case that poitns to progress you've made towards a larger goal.

So back to the money... In order to justify spend (time and money) on my social media campaigns, it seems logical to me to then connect my qualitative and quantitative findings in the following ways:
1) Plot highest levels of social media activity (buzz) with sales levels to look for patterns
2) Find a way to identify the bloggers, Twitter-ers and others who participate with my company in the social media sphere and determine if they eventually buy (No, I don't know how to do this... I will keep looking for ways to accomplish this formidable task, though.)
3) Survey my market to determine impact - poll them to determine if social media may a difference in their buying decision, ask where they heard about us, etc.

With a combination of these three, perhaps I might inch closer to a better understanding of the social media/profit equation. I have yet to find a true answer to the question, "how do you measure the effectiveness of social media?", but I do think that I'm closer to reformulating the question. Stay tuned.

Wednesday, December 17, 2008

Can trust be measured? Daryl Mather thinks so.

Related to loyalty, but not to a loyalty program, per se, Daryl Mather, Senior Reliability Consultant at Meridium wrote an interesting post about measuring trust. The article seems to target professional consultants, however the concepts apply to all types of, well, "reputation marketing" for lack of a better term.

What I love about the article is that it actually talks about REAL NUMBERS. And for me, REAL NUMBERS equates to accountability.

Mather suggests that trust can be measured in the following ways:
- % of business from repeat clients (Mather says 30% of your work should come from repeat customers)
- % of business from both passive referrals and auto-referrals, where passive referrals come from clients who refer you to others, and auto-referrals are the result of your blog, press about you, etc.
- quote strike rate - ie, how many quotes are accepted vs the number sent out

These are really simple metrics that could be used for many purposes. Substitute "quotes" for "personalized letter and brochure" and you have yourself a recipe for direct mail metrics, for example. Mather's post focuses on a client-consultant trust relationship, but this could easily be translated into a consumer-brand relationship. Essentially using Mather's metrics as a foundation, you could build a scorecard for any type of business trust relationship.

Saturday, December 13, 2008

Measuring Social Media Marketing

In doing some research for this blog, I came across a great little article, written over a year ago by Michael Brito, Senior Marketing Manager, Yahoo! for the Search Engine Journal.

It's worth a read.

Brito suggests that measuring the effectiveness of social media isn't all that complicated. If you're using social media to drive traffic to your website, then you can use traditional web marketing measurement metrics (say that three times fast) such as "unique visitors", "time on site", etc.

In terms of measuring social media "at the source", so to speak, the metrics are quite similar and essentially boil down to how many interactions are had between target audience and interface.

At the end of the article, Brito points to the idea that these metrics still have to be related back to revenue, but doesn't give any suggestions as to how this might be done. I see this as key - afterall, if we don't know how much money ouractivities are bringing in, then is there a point to participating in the whole social media sphere at all? Looks like I'll be doing some more research on this topic.

Thursday, December 11, 2008

How do you measure the value of a loyalty program?

Today I attended a roundtable presented by the Toronto Chapter of the American Marketing Association entitled "Cracking the Code: The Future of Loyalty Marketing". The panel included Steven Allmen, GM, Business Development; David Soberman, Professor of Marketing, Rotman School of Business; Pat McGoey, Vice President, Marketing Solutions, Loyalty Program Consultant; Rubina Havlin, Managing Director, Credit Cards, Scotiabank and Chris Whitaker, Senior Vice President, Partner, OSL Marketing. The panel was moderated by Jim Warrington from Fantail Communications (who is also a Past President of the AMA).

Loyalty programs have always been of interest to me, but I have only really dabbled in them in my professional life. So today's session was particularly interesting to me. I came to the session wanting to know how they work, and of course, how the success of loyalty programs may be measured.

So how do you measure a loyalty program? Is it by the sheer number of customers that sign up for a particular program? This hardly seems sufficient since as many of the panelists pointed out, many of us carry around wallets full of plastic cards from programs we've abandoned long ago. If not that, then what other variables may be measured? Purchase value? Frequency? All of these are surely worth measuring, but at the end of the day, do they mean anything? Even in combination I am not convinced that this even scratches the surface.

One of the panelists (if this is you, please feel free to let me know and I will amend my post), suggested that it comes back to "lift, shift and retention" (among other factors). I see how this is the case, but I am still baffled as to how this information could be coupled with other data as well as qualitative feedback in order to distill any clear sense of "value". What do you say to a CFO who asks what the dollar value is at the end of the day?

Sorry, folks, but I don't have any answers. If you do, I invite you to share them with me.