Copywriting can be automated (but anecdote, analogy and parallel cannot)

Automated Insights, a US startup, provides software that’s used, most notably, by Associated Press to create earnings reports and baseball game reviews.

The company itself advertises its software as turning ‘spreadsheets into stories’.

There’s no doubt this has implications for journalism. Though it may not reduce investment in staff, it is allowing publishers to cover more ground, for example lower league baseball that previously might not have merited a game report.

So, what does it mean for content marketing and copywriting? Well, if you’re creating news, press releases, or product descriptions, you better make sure your work is distinctive enough to rise above what competitors with computers can do.

Some humans are blessed with the ability to use anecdote, analogy, humour and more; in general we can answer the question of ‘why’ something is, as well as the nuts and bolts of who, what and when.

automated insights

Segmentation and personalisation can be done algorithmically (and still annoy the customer)

Recommendation algorithms have long improved conversion in ecommerce. But that doesn’t mean this form of machine learning always improves customer experience.

I recently watched a talk by Mark Killingly of the RFU; he was discussing sensitivity and nuance in communicating with customers.

Mark gave the example of Amazon’s recommendation algorithm, which continued to push Moroccan garden ornaments on him after he once bought one for his mother (who had been on holiday to Morocco).

After his umpteenth session revealed yet more Moroccan garden ornaments, Mark said Amazon nearly lost a customer.

Elsewhere, in website and email segmentation, clustering algorithms and propensity modelling can be used to optimise content and communications in the sales funnel.

However, the various content that is used for this communciation must be created by a marketing department and imbued with brand personality. That means automation makes more work for people.

The point is that machine learning only works if all the inputs are correct. Manual creation and curation is needed and also helps to mitigate for errors by virtue of the personality of your marketing department.

Machine learning only uses tangible metrics but brand awareness and consumer sentiment are hard to measure quantitively and then add back into a calculation.

So, personality is the art that acts as foil to the science of martech.

one does not simply broadcast

Advertising is now targeted and automated (but display has had its chips)

We can’t argue that automated bidding hasn’t made Google into a behemoth. PPC is a pretty boring topic because of the general acceptance of its incredible efficacy.

However, when it comes to display advertising, the IAB admits the whole industry has been derailed. Crap ads at scale; it doesn’t matter if they’re notionally targeted at a demographic.

The display ad industry is in need of more personality and creativity than ever (the IAB advocates a scaling back but that’s missing the point).

Native experiences (however you define them) are on the rise, where brands can use better formats and harness editorial staff to please the reader with content aligned to brand goals.

Companies are now community managers (and they all want your love)

This point was eloquently made in this Medium article on ‘the intimacy economy’ and focused around organisations’ attempts to ingratiate themselves into our lives.

Social media is often the platform used to do this, with the theory being that the more authentic a relationship that can be formed between organisation and consumer, the more success they will have (money they will make).

So how does having personality help here? In short, it’s a game of hawks and doves.

If every brand is a hawk (adopting the same unconditional positive regard, infantilised marketing speak, imploring people to share, comment, donate and buy), perhaps the doves will shine through (toeing a time-honoured brand line and remaining true to principles and a mission statement).

Granted, this sounds wishy washy. What does it actually mean?

Here’s one example. Remember when lots of brands implored people to ‘Like us on Facebook’ or ‘Scan QR code to Like us on Facebook’.

This was brands asking consumers to do something for absolutely no benefit, just to increase awareness of a company.

A far better use of marketing real estate is to convey the personality of your brand, for example Pret a Manger stating its green credentials on its recycled napkins.

With every marketing initiative, ask yourself ‘will this please the consumer and does it fit with the heritage of our brand?’ That’s not rocket science, but it gets close to defining what brand personality is.

like us on facebook