Two fields. One AI step. That’s all it takes to classify an account with confidence.
Industry classification is one of those deceptively messy fields. Sometimes it’s missing altogether. Other times, it’s inconsistent — “Fintech,” “Financial Technology,” “Finance.” And when you rely on that data for territory planning, account scoring, or government bids, those inconsistencies can create real friction.
NAICS codes help solve that by providing a universal classification system. But enriching accounts and assigning the right code manually takes hours, especially when you’re working with limited input data.
Now, all it takes is a company name and a domain.
One AI flow step with Complete AI enriches the account and returns a standardized NAICS code.
In this walkthrough, we’ll show you how to:
- Enrich firmographic details like industry, revenue, and employee count exclusively using a Large Language Model (LLM)
- Automatically assign a standardized NAICS code
- Return the code’s title and a confidence score, without scripts, rules, or third-party enrichment tools
Let’s break down how to enrich accounts with AI — step by step.
How to Enrich Accounts and Assign NAICS Codes with Complete AI
If you caught the video above, you saw how we used a single AI-powered flow step within Complete Leads to enrich an account record and classify it using a NAICS code — all from just a company name and domain.
In the walkthrough below, we’ll break down how the multi-field enrichment step works, how to select the right context fields, and how to return additional details like industry, revenue, and confidence score — without any third-party tools or data providers.
1. Add the Multi-Field AI Enrichment Step to your flow
This use case leverages our new multi-field AI enrichment step, which lets you enrich and update multiple fields in one go, including running AI-powered standardization or classification tasks like NAICS code assignments.
To get started, insert the multi-field enrichment step into your flow. This is where all the enrichment, classification, and field updates will happen.
2. Choose fields to provide your AI of choice with context
Next, choose which fields your selected AI should use as context when making its enrichment decisions.
In this example, we’re using ChatGPT and just two inputs:
- Account Name
- Domain
This approach means you don’t need to manually write long, detailed prompts or hardcode logic. You just select the fields you want AI to consider, and it handles the rest.
3. Select the fields you want to enrich and update
After setting your context, define the output fields you want the AI to enrich.
In this example, we’re enriching the following:
- Industry
- Annual Revenue
- Employee Count
- NAICS Code
- NAICS Title
- Confidence Level
All of this is done in one flow step with no additional logic, mapping, or scripting required.
Here’s the prompt we used in each field:
| NAICS Code Infer the company’s primary NAICS code. Prioritize signals from name, website/domain, industry, description, and address. Return one primary 6-digit NAICS. If uncertain, suggest up to 2 alternates with probabilities and mark confidence low. |
| NAICS Description Provide the official NAICS title for the chosen code, then add “Confidence Level.” followed by either High, Medium, or Low (choose one). |
4. Run the flow on a sample record
To test the setup, we run the flow on a record with just the account name (Pfizer Inc.) and domain (pfizer.com). There are no enrichment tools or external data providers connected, just the AI step doing the work.
After running the flow, the AI successfully pulls in all firmographic data, assigns a NAICS code, provides the code title, and returns a confidence score, all from just two fields.
5. Review the classification and enriched output
Once the flow runs, review your record. You’ll see:
- Clean firmographic data filled in
- A standardized NAICS code and corresponding title
- A high-confidence score telling you how reliable the classification is
This enriched and structured data can now be used downstream, whether for lead routing, segmentation, reporting, or syncing with external systems.
And while this use case focuses on enrichment and classification, the same step can support standardization too — like ensuring your industry data follows a consistent naming convention across your Salesforce org. You can even layer it into broader standardization flows like the one we covered in this blog post on phone number formatting.
Complete AI Helps You Build a Data Foundation Teams Can Trust
NAICS coding is just one piece of the puzzle. The real win comes when your team uses consistent, reliable data across Salesforce. By enriching and classifying accounts in the background, you free ops from fixing records and give them time to focus on driving revenue growth.
Complete Leads and Complete AI make account classification effortless and consistent. They embed AI-driven enrichment into your Salesforce workflows, so every record enters clean, consistent, and ready to drive your next go-to-market motion.
See how Complete Leads and Complete AI simplify enrichment and classification.
Book a demo today to put AI to work in your Salesforce flows.
