Hero image of emojis to represent sentiment
15th October 2019
Signal Editors

Top tips for sentiment analysis

There’s the age-old adage that all publicity is good publicity. We’re fairly sure that whoever came up with that probably didn’t know about sentiment analysis. Or have access to a sentiment analysis tool – because they’d definitely change their mind!

Sure, seeing your brand gaining traction in the media is great. Combined with campaigns getting column inches and spokespeople collecting quotes is often a good sign that you’re doing a great job. Reams of press cuttings can only be good, right? But it can also be a sign that you need to dig a little deeper.

What can you do to understand why you’re seeing so many mentions?

Even if you’re dealing with slightly less press than the biggest of brands, you should invest in a good sentiment analysis tool. Not only to make sure that you understand why you’re being mentioned, but to inform your PR strategy going forward.

Need some info on sentiment analysis? You’ve come to the right place.

What is sentiment analysis?

So what even is sentiment analysis?

Well, in short, it’s ‘opinion mining’. When we talk about sentiment analysis, we mean the process through which PRs professionals analyse and judge media coverage in regards to the feeling it is based on. So, the preconceived conclusions and thoughts that have inspired how an organisation or spokesperson is written or talked about.

Sentiment analysis within a media monitoring tool is a little different. In this case, the tool is categorising the conversation about your brand (or campaign, or spokespeople, etc etc) as positive, negative, or neutral.

Every mention will be tagged based on the language used – for example, check out these tweets;

PhraseSentiment
‘I absolutely love the new series of Succession!’Positive
‘I can’t stand the new series of Succession. So bad.’Negative
‘I guess the latest Succession season on HBO is ok’Neutral

But it can be tricky for tools to categorise some words and phrases correctly…

‘The new series of Succession is totally sick!’ could be deemed negative, due to the use of the word ‘sick’, which is traditionally used in a negative context. But! A common slang trait is to invert negative words, so ‘sick’ becomes ‘amazing’. Still with us?

Well, we’ve got an answer for that, but before we get to it – let’s dive into how and why sentiment analysis can help you.

CTA for download of AI revolutionising PR report

Track key topics for business

First of all, it’s important to actually focus on the topics that mean something. There’s absolutely no point in tracking tons of areas that won’t give insight into what you actually want to understand.

It’s also important to focus on areas or topics that are permanent to your business, rather than spending a lot of time on tracking micro-campaigns or social posts.

So, if you were American Airlines, and you wanted to understand the sentiment around your brand – rather than a specific campaign, for instance – you’d want to track keywords that relate to your overall brand and services;

airline, comfort, space, legroom, service, snacks

rather than campaign specific keywords;

tv ad, tim mcgraw, stand up to cancer, SU2C.

Negative customer feedback

One huge benefit of using sentiment analysis strategically is to use it to understand what your customers are saying about you – and, even more importantly, what they’d change about your products or services.

Learning that 800 people love a feature is great. Slapped backs all round! Learning that 800 people hate a feature might not feel quite so great. It might mean some overtime and disappointment. But negative feedback is a gift, and hearing it means you’re able to take it in, iterate, and improve in areas you might have missed previously.

Gaining deeper insight into the voice of the customer is one of the key use cases for sentiment analysis. As well as understanding the ebb and flow of customer sentiment towards you over time, it’s a great way to surface problems that can be fixed quickly. Allowing you to route them to the right teams.

Learning what people really think about your offerings by listening in, and improving on their pain points, is beneficial to making the business perform better.

Increase positive feedback by collaborating with thought leaders

One way of helping your sentiment dial swing into the right position is to collaborate with thought leaders in your space.

Aligning your brand with influencers who have large followings can help expose you to the audiences you want to be seen by, and if they’re trusted already, this can definitely rub off on your brand in a positive way!

Let’s say you want to be regarded as an expert in the automotive world, and your engineering team have recently shipped some tech that you’re hoping will be taken up by the likes of Tesla (it’s good to have dreams, right?).

You’re going to need to get noticed, and be spoken about in the same breath as already-trusted tech.

To get your name mentioned in the right places, you could undertake sentiment analysis on influencers in the automotive space, and see who you’d be best-off nurturing relationships with. The right influencer with the right audience can propel your brand forward – so make sure you’re collaborating with the right ones!

CTA Download for PR & Media Intelligence Glossary

Track your campaigns and PR efforts

Sentiment analysis is a really simple and quick way of tracking and measuring how your campaigns are being perceived.

As we’ve discussed before, benchmarking and measurement is really, really important. How will you know whether your brand is being perceived positively or negatively without measuring?

There’s absolutely no point creating excellent work, spending ample budget on campaigns, if you’re not going to track how this work is affecting how the market views your brand.

With a few clicks in a sentiment analysis tool, you’re able to gauge public opinion on whether your dollars are being spent well, or where you might need to have a rethink on your strategy.

If you’re seeing that you’re gaining more negative mentions than positive, that’s a good indication that you’re going to need to look into why.

Artificial Intelligence: the future of sentiment analysis

We mentioned earlier that categorising sentiment is a tricky thing.

With slang and sarcasm running rings around traditional sentiment analysis tools, you need to choose an AI-boosted, intelligent machine learning platform like Signal AI to take the heavy lifting out of sentiment analysis.

Discover why Signal AI is the most intelligent sentiment analysis tool in the market with a free demo.