How we're changing the game with predictive analytics
Most agencies build and market websites for a wide range of industries. They may be good at different marketing channels (Facebook, email, AdWords), but only get to know each client’s industry on a superficial level. In other words, they may be channel-proficient, but they’re not industry-proficient.
At Coalmarch, we’re both. We’re PPC experts, SEO experts, and content experts, but we’re also service industry specialists. And we’ve got the grass stains and bug bites to prove it.
This means we get into the weeds on industry-specific trends and tactics. For example, we always recommend after-service emails that encourage your customers to leave a review. Why? Because we know that 25% of potential customers will visit your site, leave to do some brand research (mostly looking at local reviews,) and only if they like what they see, come back to your site to contact you.
We’re always looking to take this to the next level. And lately, we’ve been working on a solution that will.
To describe it, first let me ask you a question. How do you plan for next year? Do you attempt to predict next year’s revenue? If you do, how accurate is your prediction? If you come up short or exceed your prediction, do you have a reason why?
What we’ve seen is that most companies don’t have predictions. They just have goals - an important distinction. The goals are normally based on last year’s performance, with an optimistic 10-20% growth thrown in. If revenue comes in under or over the mark, there may not be any explanation given.
We aim to change that.
When we talk to our clients about what external factors influence their business, one theme comes up the most: weather. If March is especially warm, their selling season may begin early, which can throw off the rest of the year. Best case scenario, this means you have to adjust what you planned (cash flow reports, marketing spend.) Worst case scenario, you may not have properly staffed your company - so workers may have to work overtime to make up for it, or in the event of a short season, you may have to let someone go earlier than expected.
But what if you could use the weather to predict the behavior of your customers? What if you knew that, if March was 5 degrees warmer than it was last year, you’d get an additional 12 customers?
In our research, we’ve found that there is a strong correlation between the weather and your potential customers’ interest in your services.
If you joined us at our inaugural CO2 conference last December, you may have seen me mention a tool we’re calling “LeadCast.” This tool takes three inputs that we gather for a client: historical weather data from their service area, historical Google Trends data for their main keyword, and historical traffic data from their website. Relationships are established between all three data sets so that dependencies can be determined.
Long term planning
With seasonal businesses, some months are more important than others. When the season starts in March and April, you know that you need to be as busy as possible to offset the slow months of December and January in the winter.
If you predict out the entire year, you can forecast exactly how many leads you’re likely to get in March and how slow it will be in December. If you have a relatively new website, your organic traffic may not be strong yet (this is normal,) so you may choose to invest in Google Adwords to supplement the traffic you’ll get via organic search.
If you take this forecasted traffic and calculate leads, sales, and revenue, you can add up your total forecasted revenue for the year and compare that to your goal. If you’re short of your goal, you may decide to invest more in organic search or increase your AdWords budget in the months you know you will get a positive return. As the year progresses, you can adjust your priorities, time and investments to make sure you’re on track to meet your goals.
Short term planning
Here’s the one drawback to this process: we all know that long term weather forecasts are rarely accurate. They are good for high-level planning but break down pretty quickly when you try to narrow down to a week or so in the future.
The solution? Use local forecasts for short-term planning. The 7-10 day forecasts from companies like weather.com and weatherunderground.com are very accurate. So you could, on a Friday, look at the forecast for the next week and update your plans accordingly.
If it is supposed to rain, you may get a decrease in lawn care leads but an increase in pest activity. If it’s supposed to be extra hot next week, you may want to schedule extra technicians to account for the excess service calls you will get. If you build this into a weekly process, you may be able to decrease the number of surprises you experience due to the weather.
Be a more informed business owner
Your sales staff and marketing agency are responsible for the bulk of your leads, but sometimes other factors are at play. Let’s say you finished a week and you exceeded your lead goals by 40%. Do you just move on to the next week or do you reflect and evaluate? By understanding the influence of the weather, you may look back at the week and realize that it was unseasonably hot, and half of your excess business was likely due to the heat.
By the same token, you could learn that an especially cold or rainy week caused a decrease in sales - not an underperforming sales staff.
Evaluating your performance in the context of the weather is an essential strategy to streamlining the way you market and run your business. It will strengthen your ability to manage your marketing budget, adjust lead goals, and plan daily operations.
Want to learn more about our research and analysis? Join me for a live webinar Tuesday, April 11, 2017 at 2pm! I’ll share my predictive models, getting a little more in-depth on just how LeadCast works, and take questions from the Twitterverse. Sign up below, and use the hashtag #CMLeadCast before, during, and after the webinar.