Case Study: End-to-End Consulting with AI
1. Case Study: End-to-End Consulting with AI
In this final lesson, we'll cement everything you've learned by walking through a complete research, analysis, and client delivery workflow.2. Example: B2B SaaS
Let's imagine we're consulting a B2B software-as-a-service subscription company. Growth has slowed, and leadership anticipates hidden churn risk in the form of more billing complaints, more downgrade attempts, and declining user satisfaction.3. Example: B2B SaaS Growth
We're still early in the research phase of the project. So far, we've gathered data from several departments. We have customer support tickets,4. Example: B2B SaaS Growth
voice of the customer surveys, which are sent right after key interactions in the subscription journey;5. Example: B2B SaaS Growth
and qualitative notes from direct conversations with customers.6. Example: B2B SaaS Growth
Finally, we have access to digital journey analytics - step-by-step data showing how customers move through the product, based on tracked events and sessions.7. Example: B2B SaaS Growth
To finalize the research phase, we need to bring together all of the data currently spread across multiple tables into a single, consolidated table, if possible. To do this, we can use Copilot to generate an Excel table that aggregates information from the original sources. It also clearly flags any data that cannot be aggregated, making sure nothing important is missed in the process. In a real consulting project, this would be a natural point to pause, run consistency checks, and make sure the AI system has not introduced any errors. We'll assume the analysis is valid from the off.8. Example: B2B SaaS Growth
During the analysis phase, AI systems can help by suggesting chart types that best fit our data. This makes it easier to explore the data and quickly understand the main trends. Looking at the model's response, we can see that it proposes five different charts, explaining both the chart type and the insight each one highlights.9. Example: B2B SaaS Growth
At the end, the model offers to generate the charts for us. Let's confirm that request and ask it to combine all of the charts into a single PDF, so we can easily review the analysis at any time.10. Example: B2B SaaS Growth
From the generated charts, we can see that the main drivers of dissatisfaction are "Plan downgrade blocked" and "Payment failed". Let's now ask for more targeted charts that focus specifically on these two issues.11. Example: B2B SaaS Growth
It seems clear that "Plan downgrade blocked" and "Payment failed" are the main pain points. But wait! We also have textual data that couldn't be aggregated into the main table. Let's check whether this qualitative data supports our findings with another prompt to review the interview notes and our suspected pain points. We can see that the model confirms the interview notes strongly reinforce our conclusion: "Plan downgrade blocked" and "Payment failed" are indeed the most painful issues for customers, according to the data.12. Example: B2B SaaS Growth
Finally, let's begin the client-delivery phase and ask the model to produce a one-line executive summary that we can use in a slide deck. From here, we could create a detailed narrative from the findings of the analysis, including key visuals to strengthen it.13. Let's practice!
In this video, we've illustrated a simplified consulting workflow where AI systems help speed up productivity and support complex tasks like data aggregation, chart creation, and client-ready message refinement. But remember: use AI wisely, and always double-check AI-generated insights and statements. Time for the final exercises to cement everything you've learned!Create Your Free Account
or
By continuing, you accept our Terms of Use, our Privacy Policy and that your data is stored in the USA.