The drug industry still faces occasional shortages of key products. One potential reason? Lack of access to timely data that would enable manufacturing and operations to respond more quickly to temporary demand blips, or anticipate longer-term changes. Most firms already have access to this data, but not everyone is aware that it exists. Edge Dynamics’ Chief Strategy Officer, Richard Prest, shared his views recently. Here is some of what he had to say. For the audio interview, click here.
PhM – Please comment on how the industry is managing drug distribution today. What are most companies failing to realize about what they can do to reduce supply outages?
RP – In the pharmaceutical industry, most manufacturing is built to plan rather than to built to order, so both supply and demand issues can affect whether that plan is what actually occurs. The risks are much higher on the demand side than on the supply side. Sensing changes in demand in the channel is a good way to get an early warning on demand-side risks, and to be able to respond to them in a timely fashion to minimize impact. If you see that demand for a product is higher than expected, for instance, you can adjust fill-andfinish inventory quickly and minimize shortage problems.
PhM – How far is pharma behind other industries in adjusting its manufacturing to demand? Are there any benchmarks it should adopt from other industries?
RP – It’s always hard to put an exact figure on these things, but pharma has had stronger margins than other industries and thus hasn’t focused on establishing operational excellence in this area. Now, increased margin pressure is forcing companies to squeeze every bit of revenue and profit from every operation, so the drug industry is looking to other industries for ways to reduce inventories and improve fill rates.
PhM – How about electronic data interchange (EDI)? We’ve heard for years about how it’s used in other industries. Has pharma adopted this technology to connect with distributors?
RP – When we started Edge Dynamics about six years ago, we had hoped to use a Web services transaction model, but it turned out that EDI was by far the dominant form of transaction within the pharma industry. It had adoption rates of almost 100%, among manufacturers and distributors.
PhM – How much access do most drug companies have to distributor and sales-channel data? What are the limitations?
RP– There’s pretty good access. We understand that about 150 pharmaceutical companies have access to inventory and sales data, and as well as to invoice-level EDI 867 data, which provides sales numbers from distributors to the individual outlets that are downstream in the channel. Another 100 pharma companies have access to EDI 852 data that provides aggregated sales as well as inventory and other product data. Access to this data typically comes in conjunction with distribution agreements with downstream trading partners.
PhM – That’s a significant number of companies. If the information is there, why aren’t more companies using it?
RP – The big challenge boils down to organizational silos within the typical pharmaceutical company, lack of awareness of data, and the relative newness of some of this data. The 852 data has been around for some time but may not be as interesting to marketing and sales because it doesn’t provide invoice-level data. 867 data is new, and does provide invoice-level information. People are coming to grips with how relevant it is. Outside of trade, awareness drops to 25%.
PhM – What would be needed in terms of IT to better connect sales to manufacturing and ops?
RP – The technical plumbing is relatively straightforward; the problems are on human side. If you introduce a new source of sales data, everyone will compare it to the source they already have, so there is a need to explain it and to get them to trust data.
PhM – You mentioned the problem of silos. We’ve seen that some pharmaceutical manufacturing facilities and companies are “data-rich but information-poor.” How does a company prevent this sales data from becoming just one more data management problem? How can they manage it and ensure that it is available to all potential user groups?
RP – Whenever you introduce a new source of data, people need to understand its value and its limitations. Traditional forms of data on script info is slow to arrive, often after six weeks, and suffers from extrapolations, so there can be some errors in some segments of the market it analyzes. Channel data is timely but can suffers from blinding, where some of the data is not exposed for which downstream location it is shipped to. People need to be aware of limitations and there needs to be buyin that the data will help solve specific problems. A case must be made for why it will be helpful. We find that timeliness is the key advantage to channel data. In cases such as new product introductions, end of product life, competitive actions, seasonal products, where one sees rapid changes in demand, the timeliness of this information can be particularly useful.
PhM - How does Edge’s IT platform fit in with other IT, what does it pull data from and to which plant-floor systems would it be connected?
RP - We typically fit in with the system of record for EDI data, as we look for order and pricing and forecasting data. For manufacturing, SAP is the dominant ERP system and we’re pulling factory floor data from ERP. If there’s a warehouse management system we may also be receiving that data well. Our system also connects to plant historians, such as PI.
