3D Communications AI worked with a small podcast production company (client) to pilot AI for one specific use case. The objective was to increase efficiency in their production process so they could accommodate more customers who use podcasts as part of their business operations and marketing strategy.
At the inception of our work together, the client was generating only enough revenue to cover costs. Essentially, they were working at a break-even level. However, even if they were able to bring in more business, their staff wouldn’t be able to handle the workload.
Given AI’s reputation for increasing efficiency, we examined the essential and central workflow for the podcast production use case. Because the client had been operating their business for eight years, this workflow was well-established and consistent, making it ideal for piloting AI.
The workflow consisted of a list of sequenced steps required to produce a single podcast. The list shown here was then built out to include each piece of software used in the respective steps.
An analysis of this basic information helped to determine which software in their tech stack was AI-infused. Given the rapid advancement of AI, it was quickly discovered that nearly every piece of software claimed to have some AI features. However, the client didn’t have a clear understanding of the degree to which those AI features were being used in their process, and if those AI features were effectively increasing either the efficiency or quality of the overall process.
This is where it became evident that in order to pilot AI with this particular use case, input from all those involved in the workflow would be necessary. At that time, five people on the client side were involved in the podcast production process.
For the purpose of this case study, those individuals are identified by function rather than by name. However, in an actual project, keeping names visible is a good way to “keep the human in the loop” in order to ensure a better overall outcome.
This graphic is excerpted from the client’s AI Strategy Worksheet, in which the individual contributors are aligned with their tasks shown in the chart above. As mentioned earlier, those tasks are listed in the sequence in which they are performed.
By adding assignees to each task, we identified points at which handoffs and contingencies occurred and could disrupt and decrease overall efficiency. This was an unexpected but valuable observation in that it provided a significant metric.
At this point in the process, we had these baseline metrics:
Other metrics that were used in this case study:
Identifying these baseline metrics at the beginning of the pilot project played a more important role than we anticipated, as you’ll read later in this post.
The next step was to actually pilot AI infused in the existing tech stack. As mentioned above, the entire team was asked for their input about the degree to which the SaaS programs they used were AI-infused. That step in the process was valuable in several respects:
As each task in the workflow was performed over the next couple of months, the respective members of the production staff explored the website for the software they were using. This research was relatively easy, as all software is now being infused with AI, and the technology industry has recognized how important it is to get their AI features recognized and into the hands of their users.
The lead person on this pilot project encouraged and allowed the team to invest a reasonable amount of time in experimenting with the new features. She set expectations by asking staff to share their findings with each other. This was done informally for the most part. All new discoveries were celebrated, whether they resulted in modified processes or validated existing processes.
Another contributing factor to the success of this case study relates to the metrics. Each person involved was aware of what was being measured and how. They understood the importance of these metrics in determining the value of adopting and integrating AI into the process. To counter any potential fear a contributor might have about being replaced by AI and losing their job, the project lead consistently referred to the ultimate objective: to increase efficiency in their production process so they could accommodate more customers.
This case study was conducted over a 60-day period. Metrics were recorded and compared each month in order to gauge if improvements in productivity were occurring at all and how quickly. The metrics were especially valuable as a frame that guided efforts to prioritize which AI-infused software to use, acquire, or replace, depending on the impact on existing processes. They also informed decisions about modifying and changing the sequence of certain steps in the production process, as well as assigning responsibilities in a way that made more sense.
Based on the findings after 60 days, the production staff – now fully involved in the process of adopting and integrating AI – continues to present leadership with their own ideas for building momentum and increasing productivity. The time saved in this single but critical workflow is now being invested in lead-generation activities to help fill the organization's pipeline of clients.
It is expected that the workflow for this use case will continue to be refined as it is applied to the growing workload created by additional clients. The leadership of this organization is now setting its sights on innovations that will allow it to expand its offerings beyond podcast production to include marketing and optimizing podcasting for industry verticals such as real estate, travel and hospitality, and non-profits.
This case study was designed according to the Use Case Model. Download your workbook here: