A controller pulls the day's sales from the point-of-sale system. Returns sit in another file. Voids sit in a third. Nobody has joined them, so the number on the report is wrong, and everyone in the meeting knows it.
This is the most common reason mid-sized employers hire a data person. It is also the reason many of those hires disappoint. The company thinks it needs analysis. What it needs first is access: the right people can reach clean, trusted data without a favour from someone in IT.
Why the first data hire often fails
An analyst dropped into a business with scattered data spends most of the week finding files, reconciling them and arguing about which version is correct. The charts come later, if at all. Within a year the analyst is frustrated, and the owner concludes that data hires do not pay off.
The role was never the problem. The sequence was. Before you write the job description, answer three questions:
- Where does the data you rely on live today? Count the systems.
- Who can reach it, and who has to ask someone else?
- Who decides which number is right when two sources disagree?
If you cannot answer the third question, no analyst will fix it for you.
Match the hire to the actual gap
The gap usually falls into one of four types, and each calls for a different person.
| Problem | What you see | Who to hire |
|---|---|---|
| Scattered systems | Manual exports, spreadsheets stitched together | Data engineer or integration specialist |
| Unreliable numbers | Reports that disagree, constant rework | Data quality or governance lead |
| No one can read the data | Reports exist but go unused | BI developer or analytics lead |
| Unclear rules on access | Either everything is locked down or everyone sees everything | Governance and security lead |
Most companies under 500 people need the first or second row long before the third. Hiring in the wrong order is expensive in salary and in lost patience.
What this looks like in operations, finance and sales
Data access is not an IT topic only. It shows up in the functions you already run.
Operations and manufacturing. A plant manager wants scrap, downtime and labour hours on one view. If production data lives in the floor system and labour hours sit in payroll, someone has to connect them. That is an integration job, and it often pays back faster than a new analyst.
Finance. Month-end close slows when the controller waits on exports from three systems. A finance systems or reporting specialist who can automate those pulls gives the team days back every month.
Sales. Sales leaders get pipeline reports they do not trust because the CRM is half-filled and billing data lives elsewhere. The fix is part process and part a person who owns the data model.
Governance is a hiring decision too
Making data easier to reach raises a fair worry: who should see what. Employers in both countries carry real obligations here. Canadian firms deal with PIPEDA and provincial privacy law. US firms face a patchwork of state rules and, in some sectors, federal requirements.
Someone has to own access rules, data definitions and quality standards. In a smaller company that may be a senior analyst or finance systems lead with clear authority, not a full-time governance team. Name the person. Put it in the job description. Unowned governance simply does not happen.
Skills to screen for
When you interview for these roles, look past the tool list. Tools change, and a candidate who has used one platform can usually learn another. Screen for:
- Plain explanations. Can they describe a data problem to a plant manager without jargon?
- Reconciliation instinct. Ask how they would find out why two reports disagree. Good candidates start asking questions about definitions and timing.
- Judgment on access. Ask how they decide who gets what. You want a considered answer, not "everyone" or "nobody".
- Patience with messy systems. Older ERPs and spreadsheets are normal. Candidates who have only worked in clean environments struggle.
Contract first, then permanent
If you are unsure what you need, a contract data engineer or integration specialist for 3 to 6 months is a reasonable way to find out. They can map the systems, fix the worst gaps and tell you honestly whether a permanent hire is justified, and in what form. It costs less than hiring the wrong full-time role and waiting a year to find out.
What to do next
Before you post a data role, spend an hour listing your core reports, the systems behind each and the person who trusts each number. The gaps in that list are your job description. Write the role around the biggest gap, not around a title you saw elsewhere.
If you want a second opinion on which hire fits, a Traverse recruiter can help you scope the role and find candidates who have solved this problem before.