HASHTRACKER

Social listening 3 min read

How to choose between social listening tools without regret

Feature lists are useless for comparing social listening tools. Four questions separate the ones that hold up: data source, history, counted reach and export.

Cover drawn as mentions landing over time

Every demo of social listening tools looks the same. A dashboard, a sentiment donut and a wall of mentions scrolling past.

Then you buy one and discover the questions you actually had are the ones it answers worst.

What these tools are for

A social listening tool collects public posts matching a query and presents them as a dataset rather than a feed. Volume over time, who posted, how far it travelled and what formats carried it.

That is the job. Everything else on the screen is packaging.

Tell us what you need to measure and we will tell you whether a tool, a report or neither is the right answer.

Tell us what you need to measure

The questions to ask before comparing features

Feature lists are useless for choosing, because every tool has every feature. Four questions separate them properly. Where does the data come from, official API or scraping? How far back does the history go? Is reach counted or estimated? And can you export the underlying rows?

The fourth one matters more than people expect. A tool you cannot export from is a tool whose conclusions you cannot check.

Why the data source decides everything else

A tool built on the official API works within known limits and gives you figures the platform itself reports. One built on scraping gives you whatever it managed to grab that day.

The difference shows up when you need to defend a number. "The platform reported 312,000 impressions" is a statement. "Our estimate suggests around 300,000 reach" is an opinion with decimals.

For client reporting or academic work, that distinction is the whole ballgame. Everything we deliver is gathered within the Twitter compliance rules, which is what keeps it usable in both contexts.

Real time dashboards against retrospective reports

These solve different problems and the industry sells them as one product. A real time dashboard is for noticing. Something spikes, someone gets alerted, someone responds. It is worth having if you have a team watching it.

A retrospective report is for understanding. It covers a defined period, it is complete, and it can be compared against the previous one. Most of the value in listening comes from this second mode, and most of the money goes into the first.

If nobody is watching the dashboard at 11pm, you are paying for a feature you do not use.

What sentiment analysis will and will not give you

Every tool sells sentiment. Very few explain how it works. Without a trained model you get a dictionary of positive and negative words, which breaks on irony, on negation, on slang, and breaks much harder across several languages at once. A post saying "great, another outage" gets scored positive.

Treat any sentiment figure as a rough signal, never as a metric to report on its own. If a tool cannot tell you how it classifies, the number means nothing.

When a tool is the wrong answer

Subscriptions make sense when someone uses the thing weekly. A lot of listening needs are not like that. An agency measuring one campaign per quarter, a researcher gathering a dataset once, a comms team preparing an annual review: all of them are paying twelve months for four weeks of use.

For those, a report per project costs less and answers more. Tell us what you are trying to find out and we will scope it before you sign anything.

Tell us what you are trying to find out

Choosing a subscription

Every subscription carries the same product. The only decision is how many tweets you need to analyse.

See the subscriptions