TL;DR
- Critiquing what came before is necessary, but how you do it determines whether you're "the new person bringing judgment" or "the one who came to tear everything down."
- Data is your best diplomatic ally: it's not your opinion, it's what the numbers show.
- Always separate the decision from the person. The strategy didn't work ≠ someone did their job badly.
- A 4-step framework: observe, quantify, contextualize, propose. Never skip steps.
- The 5 most common mistakes: naming names, sending written critiques without warning, changing things without consensus, raising it in someone else's meeting, being passive-aggressive.
Diagnosing a previous social media strategy means identifying, with verifiable data, what was working and what wasn't — always separating the decision from the person who made it. The 4-step framework is: observe, quantify, contextualize, and propose.
The social media strategy you've inherited probably had things that weren't working. You saw that within the first two weeks. The problem now isn't what to say — it's how to say it so the message lands, gets accepted, and you don't become the new person who came to stir everything up.
This article is for Social Media Managers who just landed in a new role, finished auditing the inherited accounts, and now face the tricky part: flagging what wasn't working without burning relationships that haven't even started yet.
We're not going to talk about how to be "diplomatic" in the sense of hiding problems. We're going to talk about how to be precise: saying exactly what the data shows, in the right tone and at the right moment, so the critique is useful and not a score-settling exercise that leaves someone resentful.
WHY DATA IS YOUR BEST ALLY
When you say "I think the previous strategy wasn't working," you're giving your opinion. Your opinion is debatable. Even more so when you're new and people with more seniority have reasons to defend what was done.
When you say "Instagram's engagement rate dropped from 3.2% to 1.4% between March and September last year, and LinkedIn lost 1,800 followers without any intentional cleanup," you're presenting a fact. Facts are harder to argue with. And when you present facts instead of opinions, three things happen:
1. You remove yourself from the center of the conversation. It's not about you saying something — it's about the data saying something. You're just putting it on the table.
2. You defuse the team's defensiveness. No one feels attacked when the one speaking is the data. But they can feel attacked when the one speaking is a person.
3. You force the conversation onto a professional plane. Data forces a response with either more data or context. Either way, the conversation becomes professional and leaves the emotional zone.
The rule: before stating a critical diagnosis, have 3 specific, verifiable data points that back it up. Without those 3 data points, wait. Keep auditing.
HOW TO SEPARATE "THE DECISION WAS BAD" FROM "THE PERSON DID THEIR JOB BADLY"
This is the most important point in this article, and the one almost no one applies well. One thing is that a strategy didn't work. Another, very different thing, is that someone did it badly. The two can overlap, but they're almost never equivalent.
Reasons why a strategy didn't work (and don't imply bad work):
- The goals were unreachable with the assigned budget.
- The strategic decision came from the top, not from the team.
- The market shifted mid-year and no one read it in time.
- The available tools didn't let you measure what really mattered.
- The organizational context (layoffs, product pivots, positioning changes) had a bigger impact than expected.
Reasons that would imply bad work:
- Lack of basic technical knowledge.
- Ignoring available data and deciding by intuition alone.
- Failure to adapt to clear market signals.
- Systematically missing deadlines.
If you make it clear that you're critiquing the decision without critiquing the person who made it, the conversation moves to much more productive ground.
🤓 TIP | A useful phrase might be:
"The 2024 strategy [+ specific data point on why it didn't work], and I can see why it was made at the time, but looking ahead to 2026 we have new information that points in a different direction."
That phrase critiques the strategy with data, acknowledges historical context without judgment, and redirects forward without getting stuck in the past. Use it as a pattern.
4-STEP FRAMEWORK: OBSERVE → QUANTIFY → CONTEXTUALIZE → PROPOSE
This is the exact order in which every critical diagnosis should appear. If you skip a step, the message weakens or becomes personal.
Step 1: Observe
Describe what's happening, without interpretation. Examples: "Instagram engagement dropped from 3.2% to 1.4% between Q1 and Q3 of 2024." / "4 posts/week were published the first six months, 2 posts/week the next six." / "TikTok was opened in June and had no posts from August to December." No adjectives, no judgments. Just facts.
Step 2: Quantify
Turn the observation into a comparable data point. "That 1.4% engagement is below the sector average (2.1%) and the top direct competitor (2.8%)." / "The drop in frequency coincides with a 40% drop in reach." / "The 4 months without TikTok posts cost ~12,000 monthly impressions." Quantifying gives you the yardstick. Without the yardstick, everything is opinion.
Step 3: Contextualize
Explain why it matters. This is where your professional judgment shows: "Engagement below the sector average means our content is competing at a disadvantage — something's off with format, hook, or frequency." / "The overlap between fewer posts and lower reach suggests the algorithm penalized inconsistency." / "Abandoning TikTok after starting is worse than never starting, because users detect the inconsistency." You can have an opinion here — but it's an opinion loaded with data.
Step 4: Propose
Always end with a proposal, not a complaint. "I propose reviewing the 20 posts with the best engagement rate and building a replicable hook framework." / "I propose going back to TikTok with a minimum viable cadence (2/week) and measuring at 60 days whether there's recovery." / "I propose running a qualitative analysis that tells us why those posts worked." Without a proposal, the critique floats. With a proposal, it becomes material for a decision.
THE 5 MISTAKES THAT BURN BRIDGES INSTANTLY
Mistake 1: Naming names. "Carlos's strategy" turns a strategic critique into a personal one — even if Carlos is no longer there. Talk about "the previous strategy," "what was done in 2024." Never name anyone.
Mistake 2: Sending critiques in writing without warning. An email saying "I've seen several problems in the previous strategy" lands badly. Any significant critique goes first through a spoken conversation, and only then — if needed — gets documented. Never the other way around.
Mistake 3: Changing things before getting the green light. Changing tone, hashtags, or starting to post on a new platform before presenting the diagnosis and getting the green light is professional suicide.
Mistake 4: Raising it in someone else's meeting. Critiquing the previous strategy in a meeting where that topic wasn't on the agenda is invasive. If you want to address it, request a specific meeting.
Mistake 5: Being passive-aggressive. "Well, sure, if that's how things were done before..." — these phrases are more damaging than a direct critique because they generate discomfort without opening a conversation. If you're going to critique, critique cleanly.
HOW TO USE QUALITATIVE ANALYSIS TO BULLETPROOF YOUR ARGUMENTS
Quantitative data ("engagement rate dropped 56%") is powerful but debatable: "yes, but it's because of X" (algorithm change, seasonality). Qualitative data (the why behind what happened) is much harder to counter.
If you can say "engagement rate dropped because, analyzing the 20 posts with the best and worst performance, the worst had generic hooks like 'happy Monday' and the best had first-person hooks — and in the last 6 months, 78% of posts were the first type," you're presenting an argument that isn't easily rebutted.
This qualitative layer is what's called content intelligence: not just knowing what happened, but why it happened. If your company doesn't have tools that provide this analysis, take note — your future proposal will likely include acquiring something like this.
In the meantime, three quick paths without a dedicated tool (but boosting the analysis with AI):
- Review the 20 posts with the highest and lowest performance: look for patterns in hook, format, tone, time.
- Comparison with your competitor analysis: what the ones doing it better are doing differently.
- Alignment with buyer persona: whether the content speaks to the right audience.
FREQUENTLY ASKED QUESTIONS
How do you tell your manager that the previous social media strategy wasn't working?
Base everything on verifiable data, not opinions. Use the framework: observe → quantify → contextualize → propose. Never name names, never critique before you have at least 3 solid data points, and always pair every critique with a concrete proposal going forward.
How long should I wait before critiquing the previous strategy?
At minimum, 2–3 weeks — the time for the audit. Critiquing before that is critiquing without an argument. More than 6 weeks starts to look like passivity. Reasonable range: week 3 to week 5.
What do I do if my manager was the one who approved the previous strategy?
Double caution. Explicitly separate the decision from the person, leave room for context, and avoid phrases that sound like "what was done before was wrong." The message should be: "With the information we have today, there's an opportunity to improve."
How should I bring up the previous strategy in a team meeting?
In a group meeting, be even more restrained. Drop the data, not the conclusions. The in-depth diagnosis happens in a one-on-one meeting with your manager.
How can I make sure my diagnosis doesn't sound like a personal attack?
Two rules: present facts, not judgments ("engagement was 1.4%" vs. "engagement was bad"), and always mention as much context as possible ("there may have been strategic reasons at the time"). That defuses 80% of defensive reactions.
