Six things to consider if you're thinking about doing interviews for content

I'm an avid consumer of podcasts and interviews, written or otherwise (and am currently really enjoying the Jamie Anderson stream, among many others). So for context I should emphasise that I love interviews – as a 'punter', that is. This article is aimed at newly-minted portals that have just got on the radar of PRs, and need to take a pragmatic look at the dazzling 'chat' opportunities that are suddenly filling their editorial inbox. I've been in this position many times by now...

It's not a coincidence that I haven't solicited so many interviews in recent years, across the mid-range portals that I've worked for. Once you get into the logistics of doing them, and start to compare the benefit they bring you over a period of time, they're an unappealing prospect — even if, like me, you love meeting people and gathering up insights on subjects that are close to your heart.

If you're Wired or Forbes, not many of these points apply. But to smaller outlets that think interviews might be a lazy solution to providing worthwhile content, here are some of the things I learned...

1: You'll need to anticipate a subject's 'old chestnuts'

This isn't their first rodeo, most of the time. You're going to need some pretty Zen techniques to wake some of your subjects up from their interview 'scripts' (particularly if they are seasoned junket veterans and/or actors).

The first and most important technique is to consume as many of the subject's previous interviews as possible. I say 'consume', because a great many of them may be on video or audio, which can mean dedicating a lot of time to this aspect of research.

However, there is no other way for you to discover in advance which of your interview subject's favourite 'war stories' get trotted out on a regular basis. And you'll be mighty disappointed to find that your 'scoop' had all the juice sucked out of it back in the early 2000s.

If you have a time-limited interview slot (usually the case), you could well be faced with the difficult choice of whether or not to let your subject's 'Old Chestnut' burn through five or more minutes of your allotted time. It might be a good idea; it lets them feel comfortable, which can be fertile ground to take the chat into less 'prepared' territory.

But also, you both know that at some point you'll have to stop all that abstract chat and start plugging the book/movie/site/etc. A lengthy old chestnut is often used by the interview subject as a kind of filibuster to prevent any genuinely discursive (i.e. risky) conversation occurring before plug-time comes round. When an interview becomes this chess-like, you're going to need some real 'game'.

2: You won't be allowed to 'scoop' the major outlets

So...you're a mid-range content portal that has been able to make a reasonable impression on the market, proportional to your resources, and now you're in the PR roster for a worthwhile interview subject. In this capacity you'll usually be filling one of a few last-minute vacant slots that some of the publications ranked above you couldn't accommodate for lack of time, or for other logistical reasons.

You'll very often be given an embargo, such as 9 am or 12 pm on a particular day. But as your finger hovers over the publish button at 11:59, with all your Twitter and other social and referral submission forms primed, you'll notice from your news feeds that the top-ranked outlets have jumped the gun by at least an hour; and you can be sure that this is by arrangement. It's effectively their 'payola' for investing in a junket. None of the PR relations damage that would apply to you is going to apply to them.

So unless you got something in the interview that they didn't, your many hours of research — and then conducting and transcribing the interview — have resulted only in SEO fulfilment for the PR. A victory for them, but representing the value of three or four lost articles for you.

3: You can get your scoop pulled even after the pre-approval process

Some years ago I was roped into interviewing a famous film star who was then, and still is, involved in one of the biggest movie franchises in the world. The interview was in support of a smaller project that they had done. Though the star's association with the movie franchise was the obvious hook that the producer hoped to exploit (with the star's agreement), my publishing the interview was to be subject to approval of the final text.

So I did the interview, and we inevitably discussed the franchise. A minor nugget about it emerged which I felt could help the interview gain traction. But, because of the pre-approval agreement, I didn't get too excited about it.

However, I was pleasantly surprised to get the final form of the interview approved without any changes.

The next day it went up, and went viral. Within two hours the producer was frantically on the phone with me; the star had been contacted by the major studio behind the franchise, casting disapproval over the spoilers revealed outside of its own PR machine. The star was 'in tears'. Could I take out the part of the interview that had the spoiler?

By this time that news had been replicated across a number of major sites. It was already out there. And the interview had been approved. My source article would now be the only one that didn't carry the news!

It may seem harsh, but in this case I left the article up as pre-approved. In 4-5 similar cases over the years, this was the only time I stuck to my guns. When a PR censor-regime fails, it shouldn't be your problem. But in considering these things, you need to balance off the prospect of working with that PR again, and consider the case on merit in terms of the favour system, goodwill, and common sense.

4: People not currently promoting anything are among the best interviewees

Some of the best interviews I ever did were arranged speculatively, with subjects who were either not promoting anything at the moment, or were not much-used to being interviewed at all.

On the negative side, the lack of a project to promote means that there's no current context for talking to that person. On the positive side, you finally have a chance to get discursive with an authority or figure of interest in some or other field — often without the customary time-restraints of the junket.

In the case of high-celebrity subjects, these interviews can happen for a number of reasons, but usually it's either because the subject (particularly if somehow contacted directly) thinks it would be enjoyable and would like the attention; or because their representative sees your request as an opportunity to stop their client taking too long a break from the public consciousness.

When contacting people who are unaccustomed to interviews, you will need to stay in touch with your own journalistic ethics, and give them a positive experience of media exposure. If they shoot themselves in the foot, give them a hand — or risk damaging your own reputation and long-term viability.

5: Transcription sucks

Transcription is, without doubt, the worst task in journalism – mind-numbing, time-consuming and frequently frustrating (in the case of sub-standard recording quality where you can spend five minutes trying to decipher one word). And you can't really do anything to make this task any more agreeable, such as listening to music — which, strangely, seems to require the same part of the brain as is occupied with transcribing.

As a tech writer I keep my eye on AI-driven solutions. Platforms such as Trint come up now and again, but always seem to fall short. Google underwrites some of the most advanced AI and machine learning systems on the planet, and still can't provide reliable cross-accent, cross-idiom, cross-speech-impediment closed-captioning on YouTube.

That said, GoogleDocs provides a speech-to-text service that others have recommended, though I'm guessing (from much experience) that it works best in English language, and with North American accents.

Journalists weigh in at this thread with other auto-transcription platforms that they have tried. My experiments with automated transcription services to date have always resulted in spending as much time on clean-up as I would have spent on transcription. In any case, the machine learning boffins are still working the problem. In the meantime, Upwork and similar gig-writing platforms are filled with offered transcription contracts.

6: Interviews are among the least-read forms of content on any site

Go to your favourite major site and count the shares and comments on practically any interview. Then compare it to the equivalent social activity on any regular article (that likely took a fraction of the investment of time and effort).

Interviews are short-form authorised biographies – they're perceived as somewhere between content and advertisement, with heavy emphasis on the latter. We do them in the hope of a scoop or similar anomaly that could make the chat bankable; but practically the entire process of arranging and authorising it is designed to constitute a plug, and mitigate against any prospect of 'unanticipated' material.

Mid-range content outlets that do commit to an interview stream generally do it because of that one-in-twenty occurrence where an interview gains real traction or provides some value that wasn't anticipated. But that's a pretty harsh ROI on the work involved, particularly for a small and very streamlined content team. For the most part, interviews only usually lead to more interviews.