The Scenario
You are a market researcher. The regulatory database submission is due next Tuesday. You have an Excel workbook with 150 company activity descriptions on the Research tab scraped from websites and LinkedIn profiles.
The database requires an OKVED2 code for every company. Without the codes, the submission fails validation.
The bad version of this week:
- Open the OKVED2 classifier on the government portal
- Read each description, search the code tree, pick a category
- Copy it back to the workbook
- Repeat for 150 rows
- Spend Tuesday morning fixing the 23 codes the portal reviewer says are wrong.
The fast version is one prompt and the codes are in the workbook before the weekend.
The Easy Way: One Prompt in SheetXAI
SheetXAI is an AI agent inside your Excel workbook that reads the description column on the Research tab, calls DaData's OKVED2 suggestion endpoint for each row, and writes the best-matching code and section name back.
Open the SheetXAI sidebar and type:
For each business activity description in column A of the Research tab, use DaData to find the best matching OKVED2 code and write the code into column B and the official OKVED2 section name into column C. If DaData returns multiple candidates, use the highest-confidence match.
SheetXAI reads the description column, calls DaData's OKVED2 classifier for each row, and writes two enriched columns back. All 150 descriptions matched to official codes in one pass.
What You Get
Two new columns on the Research tab:
- Column B — the OKVED2 code at the appropriate hierarchy level (e.g., 62.01, 47.11.1)
- Column C — the official OKVED2 section name in Russian
The codes are from the official OKVED2 tree and pass regulatory validation.
What If the Data Is Not Quite Ready
Scraped descriptions are inconsistent in length and specificity. SheetXAI handles them in the same prompt.
When some descriptions are in English and need translating first
Some companies posted their activity descriptions in English.
For each description in column A of the Research tab, if it appears to be in English, translate it to Russian first. Then use DaData to find the best OKVED2 code match and write the code into column B and the section name into column C.
When you need the full hierarchy, not just the leaf code
The database requires division and group codes alongside the full code.
For each description in column A of the Research tab, use DaData to find the best OKVED2 match. Write the division code (2 digits) into column B, the group code (4 digits) into column C, and the full leaf code into column D. Write the section name into column E.
When some descriptions are too vague to classify confidently
Short descriptions like "retail" or "IT" will not produce a reliable leaf-level match.
For each description in column A of the Research tab, use DaData to find the best OKVED2 code and write it into column B and the section name into column C. If the description is fewer than five words or confidence is low, write "NEEDS REVIEW" in column D.
When you need the classification plus a sector breakdown for the research report
The report needs a distribution of companies by OKVED2 division.
For each description in column A of the Research tab, assign the best OKVED2 code into column B and the section name into column C using DaData. Then create a new sheet called "Sector Breakdown" listing the number of companies per OKVED2 division (top-level two-digit group), sorted by count descending.
The pattern: the code assignment and the sector analysis come from the same prompt.
Try It
Get the 7-day free trial of SheetXAI and open any Excel workbook with business activity descriptions, then ask it to assign OKVED2 codes using DaData. The DaData integration is included in every SheetXAI plan. For related workflows, see how to standardize company names in Excel or the DaData in Excel overview.