A soft month is when clients read every sentence twice. The report still has to lead with the numbers, name what moved, and separate what the export proves from what you only suspect. Guessing out loud is how trust erodes.
Generate your first client report free — upload a Google Ads, Meta Ads, or GA4 CSV, get a computed scorecard, and edit the draft commentary before you share.
Open with the period and the scorecard
Start the same way you would in a good month. Name the platform and the dates. Put this period next to the prior one for spend, clicks, conversions, CPA or ROAS, and revenue if you have it. Do not open with an apology paragraph and then scramble for figures to justify it.
One clear line beats a soft landing. "November Meta Ads: spend was flat, purchases fell 22% versus October" is usable. "Things were a bit softer this month" is not.
Check the non-ad explanations first
Before you blame creative, bidding, or the algorithm, rule out changes outside the ad account. Confirm the conversion action, attribution window, and tracking tags still match last month. Ask whether site outages, price changes, inventory gaps, or a landing-page rewrite landed in the same window. Note holidays, promotions that ended, or a sales team that took more calls offline.
If the client changed the goal definition mid-month, say so in the report. A CPA jump that follows a stricter conversion count is not the same story as a CPA jump with identical tracking.
Separate findings from hypotheses
A finding is something in the export or a change you can point to on a calendar. A hypothesis is a plausible cause you have not proved. Write them in different sentences so the client can tell which is which.
Finding pattern: name the metric, the change, and the comparison period.
Hypothesis pattern: label the guess, then say what you will check next.
Illustrative example (not a benchmark): "Spend held at $4,200. Purchases fell from 38 to 28 (−26%). CPA rose from $110 to $150. Finding: the prospecting campaign absorbed 61% of spend and drove only 4 purchases. Hypothesis: colder weather cut foot traffic to the retail locations that usually convert online browsers; we will compare store hours and promo calendars before we cut prospecting further."
Use a short conversation structure
Keep the call or the email body in four beats:
- What changed, with the figures.
- What you already ruled out (tracking, site, offer).
- What you believe is most likely, labeled as a working theory.
- One decision you want next week, not a laundry list.
If volume is too thin for a clean read, say the sample is small. Prefer "too few conversions to call a winner" over a confident story built on five sales.
Example paragraphs you can adapt
Illustrative figures only—swap in the client's real export.
Efficiency drop with stable demand: "Google Ads spend was $6,100 versus $5,900 last month. Clicks rose 4%. Conversions fell from 52 to 41 (−21%), so CPA moved from $113 to $149. We paused the brand-defense brand terms for six days after a budget mis-set; that alone does not explain the full gap. Working theory: the new landing page slowed mobile checkout. Next step: restore the prior page for half of traffic and re-read CPA in seven days."
Volume drop after a pullback: "Meta spend fell from $8,000 to $5,200 by agreement. Reach and purchases both fell roughly in line with budget. CPA held within 8% of last month. This is a smaller month by design, not a sudden efficiency failure. Recommendation: keep the same creative set until spend returns, then judge results against the higher-budget baseline."
Mixed signal: "ROAS improved from 2.1x to 2.6x while revenue fell 18%. We spent less on the prospecting ad sets that usually seed the remarketing pool. Efficiency looks better on paper; top-of-funnel pressure is weaker. Recommendation: protect the remarketing budget and add a small prospecting test before calling the efficiency gain a win."
Close on one next step
End with a single ask the client can approve: restore a landing page, move budget from one campaign to another, extend the test window, or confirm tracking. Save the export and column mapping so next month's comparison is honest. Clients forgive a bad month faster when the report matches the table and the plan for the next seven days is specific.