We live in the era of Big Data – huge data sets supported by AI technology producing near real-time insights and course corrections.

Louis Gudema

AUTHOR: Louis Gudema, Founder of Bullseye Marketing

But small and mid-sized businesses (SMBs), which most MSPs are, usually are in the world of Small Data – not enough data for statistically significant results. This has major implications for how you approach your marketing.

Let me give you a few examples.

Digital marketing

I recently ran an A/B test on two ads for a client. These were the week-by-week results:

With about 30,000 impressions a week Ad A had a pretty consistent click-through rate (CTR) – about .3%. But Ad B varied a lot from week to week, almost two to one. Some weeks you would have said that Ad A was better and other weeks Ad B, although after three weeks they were pretty close.

But does a slightly higher CTR necessarily mean better? I know people who work in marketing at one of the top 10 insurance companies in the USA. They’ve told me that they get significantly different results if they advertise “auto insurance” instead of “car insurance”. The people who click are demographically different and interested in different plans.

what’s really important is not the CTR it’s whether they ultimately turn into a customer

So in this A/B test maybe Ad A was attracting a different customer than Ad B. To find that out we’d need a lot more data because what’s really important is not the CTR it’s whether they ultimately turn into a customer. And for an MSP that sales process can take months.

Hmmm, it’s complicated.

RELATED

Spitballing ROI

Grow your MSP faster

Marketing software

A few years ago, I attended a demo from a company that sells predictive analytics software. This software uses the data in your CRM and on the public internet and, based on which companies became customers in the past, identifies those companies that are most likely to become customers in the future. This can be very helpful for your marketing and salesforce.

Sometimes the correlations (and these are correlations, not causations) are pretty easy to predict: a company selling office furniture (back when we had offices) would look for companies that had just taken out a new lease or started construction on a new facility. Sometimes they are harder to predict: one company’s best prospects were companies with a significant number of new patent applications.

At the demo I asked what was the minimum amount of CRM data they needed for their analysis and they responded that the company needed to have at least 400 customers and 100,000 people in their CRM.

That’s fine for big companies, but not many MSPs will have that much data.

That’s fine for big companies, but not many MSPs will have that much data.

Direct mail

And I know one of the top marketers for a major consumer company. They market to households nationwide. He told me that to test a new postcard they send out 100,000 cards and expect a response rate of .3% (about 300 out of 100,000). Based on their results, they can then market to millions of households.

On the other hand, if your company was focusing on, for example, marketing to 500 companies, you would need to send a sample to 218 of them – almost half of your target group — to have a 95% confidence level. That doesn’t leave many to apply the results of your test on.

However, I have created some campaigns and made some marketing changes that had huge impacts.

Simple changes for better results

For example, I created a video for one client that had a 600% higher response rate than anything else they had ever done. For another client, just by adding a button that said “Learn More” to their ads their CTRs increased 10X.

A third client had a website form for signing up for demos that literally not a single person had used in a year. They were convinced that people in their market (physical corporate security) were so concerned with privacy that they would not give up their contact information to sign up for a demo. When I re-engineered the process, though, we got dozens of sign-ups in the first month.

when you get big results, even with a small sample, you shouldn’t ignore them

For small- and mid-sized companies, a 10% difference in marketing campaign results is rarely statistically significant. But when you get big results, even with a small sample, you shouldn’t ignore them.

But when one client got 3X more leads on one day than usual, when they asked “why” my honest answer was “random noise”. We had done nothing different that day, and the leads returned to their usual level the next day and most days after. These things happen.

All of these are examples of data in direct response marketing programs. However, at any moment fewer than 5% of the companies in your market are looking for a new MSP. So in the long run your best marketing investment is to build your brand and “mental availability” – that customers will think of you when they are ready to buy. At least 50% of your marketing dollars should go towards this.

your best marketing investment is to build your brand and “mental availability”

Mental availability is built by having a superior share of voice in your market. This can be achieved by emails, blog posts, video, speaking at events, putting on workshops, placing ads on LinkedIn and in professional publications – or the program of the local high school play, etc. You will find it hard to measure precisely the impact of each effort.

As the VP of Business Development for a marketing agency, we were contacted by a VP of a $750 million company that needed our services. It was our practice to ask all incoming contacts how they heard of us (something your salespeople should ask at the beginning of every incoming conversation) and when I asked, she responded: “I saw you everywhere, I figured I needed to talk with you.”

Similarly when I was doing marketing for a young B2B tech firm I created a campaign that the new head of sales was, frankly, skeptical of (but he had the good graces not to tell me that). However (he later told me) after a few months we were receiving calls from people in the companies that we wanted to sell to and they, too, were saying, “I’m seeing you folks everywhere. I figured I had to talk with you.”

This is one month’s report for that client for whom we targeted 4,000 people in a few hundred companies.

For $5,400 we generated over 250,000 brand impressions. That kind of program has a cumulative impact (“I’m seeing you folks everywhere.”) Because of that, and fine tuning our ads and placements, in that month we had 32% more clicks for 19% less money.  And we generated many leads, but we couldn’t attribute most of those leads to any one ad or campaign.

Some short-term signs that your brand programs are working are increased traffic to your website, especially increases in visits from organic search, more companies linking to your site, and – over time – more inbound leads.

So play a long game

If you’re an SMB MSP, don’t obsess over small differences in marketing results. Look for the 5X and 10X impacts. And stick to the branding efforts that may not seem to be moving the needle initially but can have a huge impact in the long run.


Louis Gudema acts as a fractional head of marketing and marketing strategist for B2B tech firms and is a regular contributor to Modern MSP.