Campaign social listening: measuring talk after an ad

Short answer

To measure social conversation after an ad, take a baseline from the two to four weeks before the campaign, compare the campaign period and the two weeks after it using the same query, and look at five measures: the number of different people writing about the brand, the share of conversation types (question, praise, complaint, comparison), the "earned" conversation left once your own posts are removed, people repeating the campaign message in their own words, and the new questions the campaign raised. A rise in mentions is not success on its own — who wrote what, and how, is what counts.

Why mention counts are not enough

When a campaign starts, brand mentions almost always rise: your own posts, paid placements, partners' shares, news sites repeating the press release. None of that is the audience's reaction. A campaign's real social effect lies in the conversation you neither paid for nor wrote: what people say to each other about it, what they ask, what they doubt. So measurement starts by removing your own voice.

A baseline and three windows

1. 1Baseline (2–4 weeks before)Normal talk about the brand: number of different authors, the share of each type, main topics.
2. 2During the campaignThe same measures; own and paid posts recorded separately.
3. 3After (2 weeks)Does talk return to the baseline or stay at a new level? Have the questions the campaign raised been answered?

Five measures

• Different authors — how much the number of people writing about the brand rose against the baseline (not posts).
• Share of conversation types — question, praise, complaint, comparison, purchase request; which did the campaign increase?
• Earned conversation — what remains after removing your own accounts, paid placements and press-release repeats.
• The message in their own words — do people restate the campaign's core idea in their own language? That shows it landed.
• New questions and blockers — "what are the terms?", "where can I get it?", "is there small print?".

Removing your own voice

Before measuring, separate three groups. First, the company's own accounts and employees' shares. Second, paid placements — influencer posts, sponsored articles. Third, news sites repeating the press release unchanged. These groups show the campaign's reach, not the reaction. What remains — posts by customers, prospects and independent authors — is the campaign's real social effect.

Separating background talk

Not all talk about the brand during a campaign is about the campaign. Ordinary complaints, questions and other topics continue in the same weeks. To isolate campaign talk, track its key words — the slogan, the offer name, the hashtag, the discount percentage — as a separate query and compare it with the rest of the brand conversation. Only posts that come in through the campaign words are campaign reaction; the rest of the rise may be general interest or coincidence.

A one-page post-campaign summary

• Different authors: baseline, during, after.
• Type shares and the main change: "the share of questions doubled — mostly about the terms".
• Three typical quotes of earned conversation, with links.
• Whether the message landed: how people restate it.
• Questions left unanswered, and two lessons for the next campaign.

Illustrative example

This is an illustrative example. A supermarket chain runs a "fresh fruit discount at weekends" campaign. At baseline, 15–20 different authors write about the brand each week. During the campaign week this rises to 60, but once own and paid posts are removed, earned conversation is 35 authors. Among conversation types, the biggest rise is in questions: "is the discount in every branch?". The message landed — people repeat "fruit is cheap at weekends" — but the branch question went unanswered. Lesson for the next campaign: list participating branches openly in the ad.

What social talk does not prove

Social conversation does not prove a campaign's effect on sales. Talk can rise while sales do not, or the reverse — many people buy but never post. Look at it together with sales, site traffic and enquiry data, and avoid cause-and-effect claims. The value of social talk lies elsewhere: it shows how the message was understood, which questions arose and which doubts were voiced — what sales figures cannot explain.

Reacting during the campaign

Measurement happens after the campaign, but monitoring runs during it too. If the same question — about the terms, say — is asked several times a day, add the answer to the website, the ad description and customer service's ready replies that same day. If negative talk about the campaign grows, check it against crisis signals; usually it means the terms were unclear, and an explanation fixes it.

Limitations

Public sources show only part of the reaction to a campaign: most people see the ad and post nothing. Closed groups and private messages are not visible. For small brands, weekly numbers may be low and move randomly. Automatic tone and type labels are a first sort; pick typical quotes by reading the text.

Common mistakes

• Not taking a baseline before the campaign.
• Counting own and paid posts as reaction.
• Presenting mention counts as a success measure.
• Not answering questions that arise during the campaign.
• Drawing sales conclusions from social talk.

Who gets the report

The one-page post-campaign summary goes to the marketing lead and to the agency that produced the campaign. Anonymise the quotes when sharing with the agency. Unanswered questions go separately to customer service and the e-commerce team, because fixing them should not wait for the next campaign. Analysing the gap between the campaign message and what comments expect in more depth is a separate job.

Measuring with Vexvon Monitoring

In Vexvon Monitoring a search can run over a date range — results are selected by publication date, so the baseline, campaign and after windows can be compared with the same query. There is an option not to count the brand's own posts: if a keyword appears only in the author's name, that post is not counted as a mention. Each result shows its type, tone and the author's likely role, and repeats are reduced to one result. More on the Vexvon Monitoring page.

Next step

Two weeks before your next campaign, start the query for the brand and campaign words, prepare a list of your own and paid accounts, and write the one-page summary template in advance. How to track a competitor's campaign is covered in competitor campaign monitoring; other articles on this topic are in the campaign, content and lead signals section. To look at it together, get in touch.