Predictive analytics is a lot like artificial intelligence. It can be a lot of hype without actual business use or solving business needs. A lot of the “old” (read current) methods work very well still. The new and shiny thing might not be good yet but will be if you give it time to mature. You can read the 3 AI vendor barriers to see its problems currently.
There is one certainty with predictive. It is a beacon towards fact-based understanding and its importance.
B2B’s Slow Adoption
In some cases, B2B is ahead and faster to move than B2C markets. Not in all and predictive is one of them and below I am listing few considerations into it.
- Low amount of data: With predictive analytics, you need a lot of data to take full use of it. The more transactions you’d have the better predictive would be to foresee and guide advertising spend. You can us it already in surveys or polling in B2B due to high amount of answers.
- Complexity of decision-making: A lot of B2B purchase decision are complex making them harder to predict or model. A lot of them don’t follow the same pattern making it hard to build a model and predict outcomes.
- Limited data: Main sources of B2B data for purchase decisions comes from CRM. Having enough data there and for it to be clean makes it another problem. You could do with your existing customers but then it doesn’t expand to new segments or industries where your product/service might fit.
Evaluating Predictive Analytics for You
Build a business case with pros and cons for your business. Go through business cases that support and discourage the purchase of it.
- Afford: Do you have the money for the investment to put into it and when do you get your investment back and start making profit out of it?
- Support: Is your organization ready to do an experiment and fail in it? Would it be extra work and are there enough people that want to do it?
- Alternatives: Is there enough value from this or is there an already existing channel or method that isn’t leveraged to its fullest. Could focusing on those first make more sense?
More Direct Approach
Three areas where predictive analytics or modelling can provide real value for marketers. Focusing on these areas where it has been proven to work will have you skip expensive modeling projects. For now.
Effective Partnership with Sales
Understand the focus and pains of sales in your organization. Making sure there is a clear understanding and communication between sales and marketing will make a difference.
- Targeting and segmentation of prospects should be clear for both
- Investment possibilities from marketing and sales
- Transparency – what is working and what hasn’t been working
Having a single source of information and data will clarify a lot of things. You can create a dashboard with Google’s Data Studio for example to make sure everyone looks at the same data.
If your KPI’s aren’t helping to grow your business in marketing and sales side, then you need re-evaluate them. Agreeing what is MQL (marketing-qualified lead) and SQL (sales-qualified lead) with both departments will make work smoother and have higher conversion rates.
Achieving cultural alignment when you are focusing on clearly defined actions and KPI’s will be easier. Everyone is working towards the same goal and they know which their responsibilities are.
Perhaps no modeling is needed at this stage of your business transformation yet?
Scoring your leads so that everyone gets value out of the lead score is a hefty challenge. Scoring is type of predictive of modeling where you try to assign future value with it to your prospect. Lead score should answer the following questions:
- Can it tell me what is going on with the lead? Time and data is needed to nail this part. Seeing in past how accounts have behaved when they were leads and what came out of them.
- How do you know the lead is good? If it’s from your target segment should be a good starting point.
- What should assign points to a lead? Interactions on your website or calls with your sales people can be good starting points when building a lead scoring framework.
Is Predictive Worth It?
Consider pursuing value closer to what you know and exhaust those almost to the limit before bringing predictive to the equation. Sometimes traditional approaches can bring more value more effectively. Especially when there are people involved. Changing how people operate inside your business takes time.
Adopting lean thinking might be another thing to consider when going forward and developing your business processes.