Over the past ~2 months we’ve met with dozens of people from companies to get feedback on our predictive modeling offering, gauge interest on what we’re building, and learn more about how users approach data-informed decision making and use causal AI.
The people/companies we’ve spoken with have ranged in size from 15-20 person series A companies to public companies with 20,000+ employees. We thought we’d cover a lot of the questions we get, either about the industry or about the product, and general themes we’re seeing from prospective customers.
Industry Questions
If we aren’t doing predictive modeling now, are we behind?
No. Organizations are just getting started with the wider adoption of accessible predictive modeling. A lot of organizations are in the process of figuring out how to incorporate it into their decision-making frameworks and are learning about how and where to incorporate it.
What are other companies doing?
This really varies from one organization to another. We’ve spoken to organizations that are essentially relying on gut feel, some using legacy style reports that don’t provide much context, and others working with their Data Science team. The companies working with their Data Science team are generally taking 8-10 weeks to get a model built, trained, and prediction run. This doesn’t include additional tweaking or scenario modeling.
Predictive Modeling / Causal AI Questions
What’s the easiest predictive modeling use case to start with?
This is really dependent on the organization, their data maturity, and the tools they’re using. I know it’s not the answer people love hearing, but it’s honest. If you’re using any sales, leads, or support management offering then you’ll definitely have a place from which to start. Also, if you’re tracking any of it in a spreadsheet, you can get started from there, and we’ve spoken to a couple of organizations that are actually doing that.
How much data do I need?
We wrote about that in a previous blog, but in short, not much. Predictive modeling or causal AI doesn’t require the same amount of data to get started as generative AI does.
How does this relate to generative AI?
We covered this in a previous blog too, if you want to read through some of the details. The short answer is that large language models and generative AI aren’t designed to do this sort of analysis. They’re two very different functions.
How do we understand the impact of a potential change?
There are statistical models you can use to actually determine how much of an impact a change may have. On that note, keep an eye on our blog and we might have more to say about this topic! This is a question we were getting often enough that we’ve put some effort into building out a solution to help address it.
Dacture Questions
Is Dacture actually no-code?
Yes. Dacture is completely no-code. We’ve built Dacture to bring predictive modeling to the people that are making the decisions. It really is just answering a few prompts. You can have your first model trained and predictions made in a matter of minutes (depending on the size of the data set).
Is data encrypted?
Yes. We use end-to-end encryption.
Can we participate in the Dacture beta?
I think this is my favorite question right now. Absolutely. The more participation in the beta, the better. If you’d like to be included, email info@dacture.com and we’ll work with you to get you early access.
If you’d like a demo of Dacture or to talk about your approach to predictive modeling, schedule time with us.
