Tuesday, September 13, 2005

Tracking Campaign Metrics

Let us start with a small quiz - you will find answers to the questions at the end of this post.

1. Assume that you are the CEO of a F500 automobile company. What would be your purpose behind running a marketing campaign ?
a) Brand Building
b) Increased Car Sales
c) Increased Leads
d) Both a) & b)

2. Assume that you are the VP Marketing of a F500 automobile company. What would be your purpose behind running a marketing campaign ?
a) Brand Building
b) Increased Car Sales
c) Increased Leads
d) Both a) & b)

3. Once you have allocated the marketing spend to various media channels, how would you judge the effectiveness of the media channel ?
a) I will select the media channel only if I am sure it will be effective
b) Effectiveness will be measured by the number of people who see the ad ( or click it if the media is online )
c) Effectiveness will be measured by the number of people who purchase cars through that channel
d) Effectiveness will be measured by the popularity of the ad for offline media and ( search engine ) ranking of the ad for online media

Now that you have selected your answers, I must tell you that the focus of this post will be on question 3 above and I am surprised how few people can actually answer the third question correctly. Think about it for a moment - if you are running an ad campaign, wouldn't you want to track all the time how well the campaign is doing ? More importantly, if your business is to sell cars, wouldn't you want to know how many cars is the campaign helping you sell ? Wouldn't you want to know how much it costs you to sell a car through the media channel ?

It sounds really counter-intuitive, but when it comes to tracking ad campaigns, a lot of marketers make the mistake of using intermediate metrics like web traffic generated or search engine ranking instead of using ROI or sales oriented metrics. A part of the problem is that tracking an ad campaign against ROI metrics is significantly harder, but for the benefits you get, it is a small price to pay.

As an example, how would you attribute credit to an online campaign for conversions that happen offline if you are not tracking your campaign on a sales oriented metric ? Well, the answer is that you would not be able to attribute sales to the online campaign, which eventually means that your metrics portray the channel as less effective than it really is. If you continue to percieve the channel's performance that way, it would not be long before you declare the channel ineffective and stop using it completely - all because you were using the incorrect metric for tracking it in the first place.

So is using a ROI or sales oriented metric the solution to all problem ? Well, as I think more about it, I realize that we haven't even touched the real problem so far. The real problem is the existence of multiple media channels, every channel so diverse that there can be no common platform to track its performance. The usage of incorrect metrics is only a way in which the problem manifests itself today.

The solution to the real problem lies in taking a holistic view of your media channels and understanding how they work in synergy with each other to help you achieve your targets. ROI or sales based metrics are really the first step in understanding that synergy. There is a long long road that lies ahead and we shall never be able to tread on it if we are scared to take the first step.

Answers:
1. Increased Car Sales. A CEO always cares about the core business and if the company is into selling cars, then the motivation behind the campaign should be to sell more cars. Yes, brand building is important, but the end goal of building a brand is again to sell more cars. A CEO would treat brand building as a side effect of achieving his target - afterall, if his cars sell more than his competitors, doesn't that give him a good brand anyway.

2. Increased Car Sales & Brand Building. Since you are VP Marketing, you would be concerned about the brand as well in addition to achieving the target sales. For you as the VP Marketing, brand building is one of the goals, not just a side effect. Infact, a VP Marketing would sometimes go to the extent of considering just brand building as their end goal, considering that a good brand would help them sell more cars then their competitors.

3. Effectiveness will be measured by the number of people who purchase cars through that channel. If you have read through the blog, you would know why.

2 comments:

Anonymous said...

I dont agree with (3).

TV might not be a sales channel for many products, but is probably a very effective marketing channel. I am sure a good ad on TV makes a big difference to actual sales.

Even if a user doesnt click an ad, he might be impressed with it, and that might have intangible effects. He might buy the same product through a different sales channel, though the deciding factor might have been a different marketing channel.

I dont think there can be any way to completely account the effects of a marketing campaign, factors like popularity of the ad campaign will always have intangible effects.

Pranav Bhasin said...

I think we are in agreement on the fact that television has a big impact on branding, but how do you know if your brand helped you achieve your sales target ? More precisely, how do you know what portion of the sales were driven by your brand ? That is a hard question to answer, and that also answers why most companies do not bother to answer this question today.

What I am saying is that every campaign must be associated with measurable metrics, so that at the end of the day, you are able to assess the effectiveness of that media channel. This becomes more important if you are using multiple channels because it tells you where to put your money next time.

As the advertising budgets are growing every year, companies are beginning to realize the importance of attributing the value to a channel and are experimenting with innovative ways to measure the impact of a passive media like television. For example, you might have noticed ads that have a small quiz at the end of the ad and ask the users to SMS the answer to a number to be eligible for a surprise gift. Getting a count of people who responded could be used as a proxy for effectiveness of the advertisement. Similarly, for things that can be purchased online, you could give out a discount coupon number in the ad and count the number of people that shop with that discount coupon.

While there is no good answer and the measures I described above are not widespread today, I believe they will gain popularity in times to come if the VP Marketing is to be able to justify his/her advertising spend.