Creating your first Data Driven Marketing strategy?
Data driven marketing isn’t as difficult or boring as you might first think. It is virtually a way of viewing marketing where you admit that the “big idea” kind of marketing we have been used too, no longer is relevant. Cause it isn’t relevant anymore. Seriously. Oh yea, you might get lucky in a campaign or two, but if there is a method through which you can be successful all the time, then why not use that instead?
Agencies build their fame on creative campaigns. They might win awards, but they more regularly base their success on how they can win the wow’s. Not from the end consumer, but from the company buying as a result of the pitch. But in a few years the pitch will be dead. Not only because I hate the pitch, but because it is one of the most worthless buying procedures in the world. Nothing in a pitch makes sense.
First off, you have agencies that over promise as they do not know the full set of pre-existing conditions at the client. Secondly you get clients that over expect as they get impressed and forget that they are dealing with human beings, not miracle makers. The alternative, being recommended or winning a client slowly, bit by bit, is a lot better. Not only because you adapt to the client more smoothly, but because you take in as much as you deliver. Winning a client becomes a learning experience. Which is also the essence of data driven marketing. You learn, adapt and improve. No need for disappointments, just deliveries.
So, enough of the politics, how do you go about running a data driven marketing campaign?
#1. Let’s start with why data driven marketing is important
Well, let’s not start with the how just yet. Let’s begin with the why as it is important to know why you should do something so radical as to base your marketing and communication on facts rather than feelings.
a. Improving delivery
Many companies send e-mails to their customers. I would say 80-90% of you do this poorly. The vast majority of e-mails I receive every day have a spam level higher than 3. This doesn’t mean they are spam, but they have such content in them that they will be classed as such by A LOT (approximately 50%) of spam filters. If you use your data sets you would know this.
If you would work with data driven marketing you would change and adapt until you would end up at levels closer to 1 which would grant you access to almost everyone on your e-mail list. Ie. today your message doesn’t even reach your end users BY DEFAULT. They don’t have the opportunity to read your communication. This is WHY you need to work with data driven marketing.
First look at the data, what it tells you and then adapt accordingly.
b. Diversifying profitability
Many companies work in verticals where profits are decreasing. There are too many similar and comparable products. It doesn’t have to be this way. There is so much data about the users out there and so many tools that simplify communication on a 1-to-1 basis, that you no longer have no excuse. Cause although your product might be generic, the way you bundle it doesn’t have to be.
On Facebook for example, there are a 600 million people who have defined themselves through likes, shares and comments on their feeds. If you work with Data driven marketing you can UTILIZE this data in order to diversify your product offering. Eg. if you are offering to sell a mobile subscription you can choose to sell it with the cheapest prize or with the exclusive Flaming banana sword in WoW.
Data driven marketing gives you the opportunity to source for WoW players in your customer base and give them the relevant offer of the Flaming banana sword, making your product A LOT different than your competitors product. This is WHY you need to work with data driven marketing so that you can increase the relevance and thus value of your ever more generic products.
c. Creating a user journey, not a destination
In the space of social web, you need to understand that shopping is not a destination. It is an experience and a journey. The definition of a target group is no longer based on demographics but on behavior. In order to grow your business you need people with different characteristics to use your online spaces. You need the activists and ambassadors, you need the haters and you need the audience who watch and simply consume.
You need to understand who is who, as you should communicate differently to either one of them. You need to know and not guess. Looking at the data you will find patterns. Not demographical patterns, but behavioral patterns. You will be able to find your different user groups and utilize them the right way. The ambassadors can be utilized to launch new products or communicative messages. The haters can be used to create dynamic discussion and buzz as well as get important feedback on your products or communication. And the consumers, or the audience shall be entertained by either great products or messages that help them consume.
This information is all available if you look at your data sets. This is WHY data driven marketing is important to you. Your user should not be treated as any consumer. He or she shall be treated the way they want to and then be led on a journey, with you, to a better and brighter future.
Those are three important, but far from the only, reasons as to why you should adopt a data driven marketing strategy.
#2. So, how do you go about creating a data driven marketing strategy?
As with any strategy, you need to start somewhere. Usually you start with your why, then you have a look at what the rest of the industry is doing and then you make strategic decisions about what you should do, then you set some kind of goals. Then you create an organization that will carry your strategic decisions into practice, you set a budget and you start delivering tactical plans.
Creating a data driven strategy is not far from this only that you turn to the available data rather than your gut feeling. Important steps in creating a data driven marketing strategy is to create an organization and find resources that will
- Determine what data assets you have available (analytics, monitoring tools, application builds, CRMs, data bases etc.)
- Create goals depending upon what your data tells you (conversion rates, member counts, buyers, downloads etc. etc.)
- Create a hypothesis that you think might lead to achieving the goals (eg. by making the button red, we will increase conversion rate by 3%)
- Test your hypothesis (no one knows, just test)
- Adapt your tactics (change according to results rather than gut feeling or taste)
It is not the case that data driven marketing takes away the gut from the equation. The gut will always be there to create the hypothesis. Any first launch of a campaign is in fact a hypothesis. The problem with todays marketing is that people create ONE BIG IDEA that shall carry the whole campaign. If that idea doesn’t work, they do not construct their campaigns to adapt, and thus they fail.
Data driven marketing allows you to create a concept that is creatively challenging and great, if and only if, you are ready to change everything about it if the data tells you to. If you are not reaching the goals, then that is OK, if and only if you are willing to adapt your tactics, phrasing, main plot or what ever it is your campaign builds upon.
#3. How to set your goals for your data driven marketing strategy
Goals are goals. They are a guess work and will always be a guess work. We try to make goals into a science by giving them names such as KPIs (Key performance indicators) and assign values to them that are understandable such as “volumes” or “numbers”. I believe this is the wrong way of defining a KPI. A KPI for me should always be a RATE of change. Ie. the derivative of any function.
This sounds like math, and it is math. It is not what normal marketing people do, but it is what data driven marketeers do. We need to understand rates of change so that we can understand where our campaign or communications strategy is heading.
Think of it this way. If you set your KPI – remember Key Performance INDICATOR – to be a number such as 2 million views on YouTube. Then you launch a campaign. You start off with 250 000 views. All of a sudden you have 500 000 views. Three months later you have 510 000 views. Your campaign is seen as a failure and you have a difficult time explaining in terms of your KPIs what actually went wrong.
If you make this same goal a “rate of change” you will much better be able to analyze not only the result, but also the tactic of your campaign whilst it is active. Our video views grew organically by 1% per month before the campaign was launched. We launched the campaign and are currently experiencing a growth rate of 5% every day. In order to reach our expected growth of 800% in a month we need to increase our “rate of change” to at least 35% on a daily basis. (all fictive and non-coherent numbers).
As you analyze your numbers you should look at how they are accelerating or decelerating. This is done by deriving the derivative and thus get the speed and direction of how the slope is turning. Ie. you get insights as to whether or not your campaign is building momentum or if it is gradually decreasing in momentum. Remember, you might still show positive results, but the results might be less and less positive every day. A number such as a view count in itself will never be able to answer those kinds of questions. However a rate of change will.
That is why it is important to define your KPIs such as rates of change.
Eg. We believe that our client report is the most important asset to our users. In 2012 we hence want to increase the download rate of the client report by 35% daily.
Booohooo… now you cry… this is boring, difficult and takes a lot of time from being creative. BUT, if you do your job, you will be able to do a lot more creative stuff in the coming steps and be more free to say whether or not your campaign was successful. Plus, don’t worry, there will be tools doing this shortly. Recommended changes as a result of previous performances. So you will only have to work towards those targets in the future and not set them yourself.
I know. It is hateful and it is a pain in the ass. But this guess work will be your best friend when it comes to extracting higher budgets for your clients as well. You just need to get them into the boat about this and argue that everything is possible if you are allowed to work towards the goals, run your first attempt as a hypothesis, then adapt your cam
#4. Don’t decide what to do, but what to do first
Use your gut to decide what first to say. Ie. from your benchmark studies, from the available data of user behavior that you have on your Facebook activity, your website, your Twitter and in your press room, use your gut to set out a Hypothesis on what to do next. Do NOT build a creative concept that will be the “end game of all games”, but decide on a creative concept that will take you one baby step further down the line.
a. Creativity and data
I think I have to discuss creativity in some more detail here in order to make sense. The current creative industry works with big ideas or solutions. In the case of data driven marketing strategies you have to think in terms of journeys or processes rather than campaigns. There is no such thing as a solution, or being done. There is simply a journey that will take you somewhere.
Your creativity should be focused on making the journey interesting. If you are not making it interesting enough you change. There is no room for pride or darlings in data driven marketing. You have to be willing to kill anything in your way to get the results you desire. You ALSO have to be able to take on old concepts and changes that didn’t previously work, if the data tells you that it is now time to use them again.
Data driven creativity is all about doing the small changes rather than finding the big ideas.
b. Communicating through data
One way of being “baby creative” is by thinking in terms of how to use changes in the data for communication purposes. Let’s use Loyalty as an example. If you get 10 000 members of a page, all users who were early in liking the page will feel rewarded if they know they were the first there. Being creative with this kind of data, your copy simply becomes “We want to celebrate the first 100 of our 10 000 members for believing us in the early days“.
This is a small thing, but a great example of how to use data creatively. Another example can be to reward every 5th customer with a certain behavior. “We reward every 5th customer with more than 100 USD in purchases this year“.
This is not creativity you might think? Well it is, only that it is creativity based on data. It is small, step by step approach, to communication. Yes there is a space for big campaigns still, just like a Sale in a store, but the day to day sales is more important in a store, just as the day to day communication is more important in the digital space.
c. Do anything your gut tells you
This is probably the most difficult thing to convey to a client. Well, except that there isn’t the need of a “big concept”. It doesn’t matter where you start as long as you start somewhere. What matters is the rate at which you move towards improving yourself. If you constantly improve yourself at a high rate, you will become more successful. You don’t need an ending, but you need a process. Do anything, change and implement. Change again and again. If you see improvement in the data, then adapt.
#5. Implement your Analytics
Now to some basics. Once you have set your goals, you have decided where to begin changing, then you should implement your analytics so that you can measure whether or not you are moving in the right direction. However, you should not only implement a metrics system. We as people are generally lazy and if you only implement analytics you will end up not checking in on them regularly enough. You should at least:
a. Set up your target pages metrics installation
This means on all pages. Your Facebook page as well as your landing page and start page of your website. You should have an end in mind, meaning that every page should have a desired goal attached to it. If that goal is a sales or if it is the click on a link or a desired time on the page, doesn’t matter. It should be implemented in the analytics tool and then continuously improved.
b. Set up reporting
You need to setup relevant reporting. This means that you should set up funnels from each traffic source. Yes, you can have bulk sinks that is called Other, but you must define your most important sources of traffic and then set up reporting thereafter. The dashboards shall NOT, I repeat, SHALL NOT, show you some fancy graphics with arrows if that is not EXACTLY what you need to do your job. Your reporting needs to be simple and give you answers if you are moving in a positive direction or not.
c. Set up notifications
You need notifications. If there is a drop or if there is some kind of unexpected, huge change, you need to be notified. This is for your comfort only, and most services offer this as a built in plugin. If you reach x or y rate of goals, then you will be notified.
d. Check up
Before you make a change, you need to have a procedure in place of how to check up on things and how to make changes so that you see if your changes made any improvement or not. Do not change before you know you can measure the rate of change.
Your strategy needs to deal with these issues just like they deal with an organization under normal circumstances.
#6. Optimize to meet your goals
Once you have your data sets in place you shall optimize along the way. Change copy, change images, change user interface, change your call to action, change your offer. Change, change, change, change until your poor head turns blue and your bank account green.
- Changing your tactics – all the time
- Changing your graphics – all the time
- Changing your copy – all the time
- Changing your product value proposition – all the time
So much for the change. You also need a space in your strategy to take care of such things as:
- How to identify important users and what to do with them
- How to diversify your product if the data gives you the opportunity to
- Where you want to spend next christmas (yeah… you read this far you can take some nonsense)
But I will cover those subjects in other posts later on as they are all chapters in my coming book. I hope this post has inspired you in some way to have a look at data driven marketing. If not, then forget about it and look yourself in the mirror, cause that expensive creme you are using will be the last bucket you’ll afford. If you don’t start grasping this, you will get run over.