The Impossible(?) job – adjusting 2020 forecasts for established brands

 

We have no recent precedents to go on, no established techniques to use, and yet the powers that be are desperate to know how the lockdown will affect this year’s numbers – “What’s the new forecast for 2020?”. For those of you burdened with this challenging task, here are some suggestions that might help.

We’ve given quite a lot of thought to this at CSL in the past few weeks and discussed it at our BI Leader and BI Expert industry forums too. In this article we consider two major factors that might affect our forecast: 

  • March and April data – how do we interpret the peaks and what may come after
  • Assessing the impact on market dynamics 

We saw sales jump up in March across the board. Supply chains where never disrupted to any great extent for most products, and pharmacies have continued to be able to order stock regularly, so it may be reasonable to assume that the increases were a result of patients being given larger repeat prescriptions (say 3 months instead of 1) in order to prevent them from making unnecessary trips out of the house during a lockdown of unknown length. 

The NHS Prescribing data can be used to assess the extent to which this is true. For your market of interest, look at the ratio of Quantity/Items to see the average amount of product per prescription. By comparing this figure in March and April to the previous average, we can assess the extent to which patients have increased “stock” at home.

Knowing this will allow us to estimate the extent of a dip to come, as patients use up that stock. 

With fewer outpatient and GP appointments having taken place in the Spring, and many of those that did go ahead being via phone call, many of us are expecting this to have an impact HCP’s propensity to change patient medication over the period. It seems reasonable to expect a suppression of the normal market dynamics. This is likely to benefit declining brands as fewer patients are switched out, but penalise growing brands with fewer opportunities to be switched in. 

Again, we should be able to use the NHS prescribing data to assess the level to which dynamics have been supressed in each market. We can review historic months for our chosen market and sum up the market share increases of all gaining brands each month (ignore stagnant or declining brands). Divide this by 100 to get a Churn rate. We can now compare the Churn in March / April to the average rate before the lockdown began.  

If the Churn rate is considerably lower, you can adjust your forecast to account for this. If the market was already fairly stagnant, there may have been little impact. 

Best of luck! Let us know if we can help.