Point out business problems so your team sees them too
Short answer Show, don't tell. Put the number next to the goal, ask what your team sees, and let them name the cause before you add yours. Feedback aimed at the person can make work worse; questions keep attention on the task. End with an owner, a date, a number to watch. For safety or ethics problems, say it plainly.
Want help applying this to your business? Talk it through on a free 30-minute call.
Book a callHow do I point out a problem without making people defensive?
Hand over the evidence and a question, not the verdict. When a person names the problem, it's their problem to fix. When you name it, it's your opinion to argue with.
| Step | What you do | What you say |
|---|---|---|
| 1. Pick one number | Choose the figure that shows the problem: missed calls, cost per lead, redo jobs, reviews | Nothing yet |
| 2. Put it next to the goal | Show this month, last month and the target side by side | "Here's where we are and where we said we'd be." |
| 3. Ask, then wait | Ask an open question and stay quiet for a while | "What do you see here?" |
| 4. Ask for the cause | Let them explain before you do | "What do you think is driving that?" |
| 5. Add what you know | Only after they've spoken, and only what they missed | "One more thing I noticed..." |
| 6. Agree on the fix | One owner, one date, one number to check | "Who takes it, and when do we look again?" |
If two honest questions don't get there, say what you see in plain words. Questions are a way to start, not a game where they guess the answer you're hiding.
Why does telling people what's wrong backfire?
Because feedback often backfires, and it does more when it lands on the person, not the work. A large review of feedback research by Kluger and DeNisi (Psychological Bulletin, 1996) pooled 607 measured results. Feedback helped on average, but the authors found:
"over 1/3 of the FIs decreased performance" (Kluger & DeNisi, 1996)
FIs means feedback interventions: any time someone is told how they did. The same paper says feedback works less well as attention moves "closer to the self and away from the task." "Your follow-up is sloppy" points at the self. "Here are 40 quotes with no reply after 3 days" points at the task.
What questions should I ask?
Borrow the U.S. Army's after action review, a short meeting held after a job to learn from it. Its field manual describes the review as one that lets people "discover for themselves what happened during the training and why" (FM 25-101, Appendix G). It also names the trap:
"The most difficult task for an AAR leader is to avoid turning the discussion into a critique or lecture." (FM 25-101, Appendix G)
The manual tells the leader to start by only asking questions, and to walk through what was planned, what happened, why, and what to sustain or improve. The Army has since published a newer circular on these reviews, TC 7-0.1 (Army.mil, February 2025). As four questions, that works for any business:
- What was supposed to happen? (the goal, the quote, the schedule)
- What actually happened? (the number, the job, the customer's words)
- Why was there a gap?
- What do we keep, and what do we change next time?
Ask about the work, never "Why did you...?" That question puts the person on trial.
What makes people willing to admit a problem?
They have to believe nobody gets punished for saying it. Researchers call this psychological safety. Amy Edmondson's study of 51 work teams defined it as:
"a shared belief held by members of a team that the team is safe for interpersonal risk taking" (Edmondson, Administrative Science Quarterly, 1999)
In that study, teams with more of it showed more learning behavior, such as seeking feedback, asking for help, talking about errors and experimenting. Google's own team research found it mattered most of five team traits:
"This was the most important dynamic in an effective team." (Think with Google, 2023)
Google also reported that its sales teams rated high on it beat their sales targets by 17%, while teams rated low fell short by up to 19% (Think with Google, 2023). That's one company's internal data, not a promise.
What builds it, in practice:
- Name your own mistakes first, out loud, with the number.
- Thank the person who brings bad news before you react to the news.
- Review on a schedule (every Monday, after every big job), so a review never means someone is in trouble.
When should I just say it?
Some problems don't wait for discovery. Be direct when:
- Someone could get hurt, or a law or rule is being broken.
- A customer is being harmed right now.
- You already asked, they agreed, and nothing changed.
- They ask you straight out what you think.
Direct still means about the work: what happened, what it cost, what must change, by when.
How do I turn "we see it" into a fix?
Write it down before the meeting ends. A realization with no owner fades by next week.
| Item | Example |
|---|---|
| The problem, as a number | 40% of after-hours calls go to voicemail |
| The cause they named | No one checks voicemail before 10 a.m. |
| The change | Return voicemails by 9 a.m.; turn on text-back |
| Owner | Front desk lead |
| Check date | Two weeks from today |
| Number to watch | Share of voicemails answered same morning |
At the check date, open the same report. If the number moved, say so. If not, run the four questions again.
Example
A clinic's missed calls
The owner of a 6-person clinic sees 120 missed calls last month. Instead of "You're missing calls," she shows the phone report: missed calls by hour. The front desk spots that 70 came during the 12–1 p.m. lunch break, when both staff step away. They stagger lunches. Next month: 45 missed calls.
A contractor's ad cost
A remodeler's ads cost $3,000 a month and brought 20 leads, $150 each, against a $90 target. He asks his office manager to sort leads by search term. She finds a third came from "DIY" searches that never book. She adds them as negative keywords. The cause and the fix were hers, so she watches the number.
A cleaning crew's redo jobs
A cleaning company owner counts 12 redo visits in a month, about $1,200 of unpaid work. At the Monday review he asks, "What did the customer expect, and what did we do?" The crew notices most redos were move-out cleans with no checklist. They write one.
FAQ
Isn't asking questions just manipulation?
Not if the questions are real and you share the facts first. Hiding what you know to steer people is manipulation.
What if they see the data and still don't see a problem?
Ask what they think the target should be. Disagree on the goal? Settle that first. Agree and still no action? Be direct.
Does this work with a vendor or agency?
Yes. Send the report, ask what they see and what they'll change, and agree on the number to judge next month.
How often should we review?
Short and regular beats long and rare. A 15-minute weekly look at three numbers, plus a review after each big job, keeps problems small.
What if I'm the one causing the problem?
Run the four questions on your own decisions, out loud, in front of the team. It shows reviews aren't about blame.
Different for you?
- The problem is in your marketing numbers: learn which ones to read first in Is my SEO working?
- Ads run but the phone doesn't ring: start with Google Ads not getting calls.
- You suspect wasted ad spend: find it in the search terms report.
- You want one screen your team can check each week: build a free dashboard with Looker Studio.
- Leads arrive but don't book: see how a Conversion agent handles follow-up.