Anthony Capon est directeur produit des plateformes numériques de RMC et BFM. Il revient ici sur un projet emblématique lancé en mars 2024 : le développement, en interne, d’un feed vidéo vertical sur leurs plateformes. Un format pensé pour répondre à trois enjeux structurants pour le groupe : engager davantage les utilisateurs, attirer une audience plus jeune, et créer de nouvelles opportunités de monétisation.

Côté dev, ça a été sport.

Notre player vidéo, exploité à la base pour des usages 16/9 n’était clairement pas prêt pour encaisser un scroll vertical rapide, fluide et sans friction. Il a fallu soulever le capot et travailler sur quelques développements clés : le pré-chargement des vidéos (critique pour l’expérience, surtout en 4G ou en zone grise), les transitions entre vidéos, les gestes natifs etc.

Le préchargement a été particulièrement complexe, nos devs ont dû ruser pour arriver à une expérience vraiment fluide. Et tout ça avec notre player, qui n’est pas le plus souple à manipuler, mais qui porte énormément de choses (analytics, pub etc.).

Résultat : une vraie prouesse de notre équipe tech interne qui a réussi à faire tourner ce nouveau format aussi bien sur web que sur app, malgré toutes les contraintes (notamment celles d’iOS et Android). Grosse phase de QA derrière pour s’assurer que ça passe partout — vieil Android comme iPhone récent — sans glitch ni lag.

*glitch: anomalie visuelle pouvant se présenter de plusieurs façons, dans le cas présent lors du scroll dans le feed vertical ou à la lecture d’une vidéo. Exemple : défilement incontrôlé du feed, lecture d’une vidéo saccadé, désynchronisation de l’audio et de la vidéo etc. 

Derrière le feed : notre routine pour produire, éditer, publier chaque jour

Une fois le feed en place, encore fallait-il l’alimenter. Et pas à moitié !

Notre flux repose aujourd’hui sur une chaîne bien structurée :

• Les vidéos verticales sont montées sous Adobe Premiere ou “découpées” depuis le live vertical de BFM (présent dans l’application BFM, dans la page du « DIRECT VIDEO)
• Titrées, taguées puis envoyées via Wildmoka…

• puis envoyées dans notre CMS maison Cléo (ainsi que sur les plateformes sociales) ****pour que nos rédactions puissent les éditorialiser, hiérarchiser, et ajuster la temporalité de publication.

Le tout repose sur une logique de curation “chaude” : les actus du jour, les formats clés des émissions, des contenus plus “froids” avec un pas de côté vs l’actualité…

“Penser ROI Day 1” : ce qu’on a mis en place pour monétiser sans dégrader l’expérience

L’une des choses que l’on a bien faites, c’est d’avoir intégré la monétisation dès la conception et la réflexion avec toutes les parties prenantes, avec toujours la même vision : développer et maîtriser en interne l’architecture et la structure technique/publicitaire de nos produits.

Aujourd’hui, le feed peut accueillir :

• des formats sponsorisés (éditos natifs ou capsules habillées),
• des OPS sur mesure,
• des formats de monétisation vidéo plus “classiques” (en cours de déploiement sur web et apps)

On a mis en place un système de règles défini conjointement avec notre régie et piloté côté éditeur : fréquence d’apparition, durée max, contexte d’affichage.

Et surtout, on a intégré la même logique que pour les vidéos : le pré-chargement des publicités pour éviter les latences et ne pas “casser” le scroll utilisateur. Pas parfait à 100%, mais on s’en rapproche…

Le résultat : une nouvelle offre avec du potentiel et, on espère, de super KPIs pour notre régie, des formats publicitaires intégrés proprement et n’affectant pas l’expérience utilisateur.

18 mois plus tard : ce que l’on a appris

Cette offre a été lancée en mars 2024 et on est très content des chiffres et usages atteints.

Pour en partager quelques-uns :

• En moyenne, le feed vertical génère 2,5 M de vidéos vues par mois
• Sur chaque initialisation du feed vertical, on observe en moyenne 7 à 10 vidéos vues
• Un temps passé en hausse constante et répondant aussi aux enjeux business

Mais on sait qu’on peut aller encore plus loin, c’est important de partager aussi ce qu’on aurait dû mieux faire ou ce que l’on doit continuer de travailler :

• Développer notre “go to market” : travailler la notoriété et la compréhension de cette nouvelle offre auprès de nos utilisateurs, notamment in-app (trop discrète aujourd’hui, le slider de vidéos verticales est situé au milieu de la HP).

• Le travail avec nos rédactions autour de la curation et de l’offre éditoriale, qui pourra passer à l’avenir par des “capsules originales”, des collections thématiques ou autres comme on a pu le voir plutôt chez des confrères américains (notamment au Washington Post et CNN durant les élections US).
• L’étendre plus largement aux autres marques du groupe, car le potentiel est important : le player vertical est déjà disponible sur nos sites et applications BFM Business, RMC et RMC Sport, mais l’offre reste à structurer.

Aligner produit, tech et édito : notre méthode

Au-delà des KPIs, c’est l’alignement produit, tech, édito et business dès le départ qui a fait la différence. Voici comment on s’est y pris en terme d’orgnisation :

• Une équipe pilote, resserrée (5 personnes), avec un·e référent·e par métiers : PM, lead dev, lead édito, sponsor business (AdTech), et un profil Marketing (sur les sujets de mesure et de marketing de l’offre).
• Rituels clairs : point hebdo structuré (avancées, blocages, arbitrages, validations), + ateliers en parallèle sur les expertises métiers (Discovery Produit, Design UX/UI, Edito, Monétisation etc.).
• Pilotage partagé :
  • Le PM cadrait la roadmap, les enjeux et la coordination.
  • Le lead tech intervenait dès la conception pour faire les bons choix tech au service des autres enjeux
  • L’édito travaillait à la pertinence de la sélection de vidéos disponibles dans le feed + s’assurait de la qualité/réactivité/agilité des outils back pour faire sa curation/éditorialisation.
  • Le business apportait les insights commerciaux et les enjeux de monétisation sur le volet tech (KPI de visibilité, UX Pub etc.)

Dans le cadre d'une réorganisation du département Marketing du New York Times, Perrine Pavageau a pris le rôle de “Program Manager”, un rôle à la croisée de la vision stratégique et de l’exécution (lire notre article ici). Son principal objectif : mettre en place des mécanismes opérationnels pour assurer le succès des stratégies marketing. On peut identifier le Program Management à la fonction d’un PMO (Project Management Officer), vu plus fréquemment dans des Directions Produit ou Tech, qui apporte aux équipes les méthodes et outils qui leur permettront de réaliser leurs objectifs. Un PMO peut aussi gérer ses propres projets, dont les acteurs ont souvent des objectifs très différents les uns des autres. Une fonction qui répond aux questions “comment on fait ?”. Un interlocuteur vers lequel les équipes peuvent se tourner pour exprimer leurs challenges organisationnels. Dans cette série d’articles, Perrine raconte 5 projets concrets qu'elle a menés dans cette fonction pour le New York Times. Aujourd'hui, l'épisode 4 : le projet "Speed & Agility".

Les objectifs :

• Compresser le temps d’exécution pour permettre un rythme plus soutenu de campagnes
• Trouver des solutions pour absorber une demande croissante de livrables à tous les niveaux (briefs créatifs, plans médias, versioning des assets pour nouveaux segments et nouveaux formats, personnalisation, A/B tests)
• Permettre plus de flexibilité pour changer les messages, les visuels ou les offres en cas de revirement stratégique.

Le rôle et les livrables du Program Manager dans ce projet au nom de code intriguant :

• Consulter les équipes sur les points bloquants et pistes d’amélioration
• Proposer un plan d’action au management et définir des KPIs
• Faire valider le plan d’action
• Suivre l’implémentation du plan d’action et assister les équipes dans le quotidien après rollout.

Plusieurs pistes ont été identifiées, dont certaines déjà évoquées dans les articles précédents. Je vais vous parler ici de la planification trimestrielle des campagnes abonnements.

Voici les “pain points” récurrents qui ralentissait les projets :

• De trop nombreux projets étaient lancés en même temps, les briefs créatifs ne pouvaient pas être prêts à temps
• La création des plans médias en interne ajoutait plusieurs semaines aux plannings
• Le principe de validation des visuels et textes créait des blocages.

La solution identifiée : anticiper, sur les campagnes d’abonnement promotionnelles, la planification des campagnes au niveau trimestriel.

Comment ?

• En demandant aux Marketing Managers de définir une stratégie en amont pour le trimestre
• En mettant en place des marketing kickoffs et des briefs créatifs trimestriels afin que les équipes créatives puissent avoir du temps pour planifier les concepts et la cadence des visuels
• En créant un système de templates ou en réutilisant des créations pour les campagnes récurrentes
• En assurant un alignement sur les deadlines pour les plans médias et “master docs” (fichier avec liste des livrables).

L’idée derrière cela : avoir un cadre de campagne déjà prêt, pour pouvoir laisser la capacité aux équipes de “pivoter”, c’est-à-dire changer de messages ou de visuels en fonction de changements de stratégies ou de nouvelles opportunités identifiées.

Un système avec des avis pour et des contre

Pour :

• Plus de temps pour les équipes créatives et média pour penser et créer des visuels et des supports de grande qualité
• Une meilleure cohérence dans la cadence de ce que voient les clients potentiels
• De beaux systèmes créatifs mieux exploités car utilisés plusieurs fois, donnant une identité forte et cohérente
• Moins de réunions car un seul kickoff et briefing pour le trimestre.

Contre :

• Le risque de prévoir une superbe campagne qui ne verra pas le jour ou sera modifiée car le contexte (politique, social ou autre) aura changé entre le moment de la conception et le moment de la parution
• En réalité, de nombreux check-ins sont nécessaires pour assurer la pertinence des plans médias et des concepts.

Selon moi, c’est une stratégie envisageable pour un type de campagne précis, avec une certaine récurrence et une équipe expérimentée dans son domaine. A ne pas envisager sur un scope multi-produit ou pour l’introduction d’un nouveau bundle par exemple.

Justement, passer sur un modèle d’anticipation pour des campagnes plus “routinières” permet une plus grande innovation lors de lancements plus spécifiques. On peut même envisager que des équipes gérant des campagnes déjà “sur les rails” puissent venir en aide à des équipes plus petites en plein lancement produit.

3 clicks to cancel: what’s changing in France?

A new regulation was recently introduced concerning the termination of contracts in France. Specifically, the « termination in 3-clicks » law requires that we should always be able to end a contract 1) online and 2) in 3-clicks or less. This applies to any business to consumer contract, including electricity, gas, mobile and internet providers, transport companies, gym memberships and subscriptions to any media or press brand.

The law outlines the user cancellation process, click-by-click:

1st click: « Firstly, the consumer clicks on the « Cancel my contract ». This button must be easy to find and understand. Any other wording used for this function mustn’t include any ambiguity. »

2nd click: “After this step, the consumer is taken to a page summarizing their terminated contract. They should be able to verify the information and, if needed, make any necessary changes. »

3rd click: “Finally, the consumer must be able to confirm the cancellation by clicking on a button or link « Confirm contract termination »… this must be directly accessible from the page summarizing the terminated contract.

The company should then confirm the reception of this cancellation and specify the date from which the contract is no longer valid. »

How have publishers in France been responding to this legislation?

Reginald Costa, Product Manager at Mediapart, an independent French investigative online newspaper, kindly reached out to share their experience on building an unsubscription journey that both abides by the restrictions in France and supports high levels of retention:

We had proposed a simple unsubscription journey as early as the end of 2016, only to backtrack, thinking we weren’t equipped enough to proactively monitor cancellations. And also out of fear of uncontrollable waves of online cancellations.

At the start of 2019, when we redesigned our tunnels, we came back to this issue. Having greatly simplified and shortened our subscription path, we didn’t think it fair not to offer an equally simple and short cancellation path.

Especially since, during user tests, some non-subscribers cited the alleged complexity of unsubscribing as a reason for not subscribing.

We took the gamble that the user experience was paramount, and that a subscriber who can leave without friction and frustration will be more likely to return. This feature has been marketed to as strong, reassuring argument in the subscription process – we almost always write « cancel online at any time ».

And so far it’s been a winning bet at every level, including for our subscriber customer service lines, which used to receive numerous calls and emails asking how to cancel.

Reginald Costa, Product Manager at Mediapart

The user unsubscription flow on Mediapart:

Click-to-cancel in the United States

There’s been a lot of talk about « Click-to-cancel » in the states, with the Federal Trade Commission proposing a formal rule in March to offer straightforward self-service subscription cancellation:

Specifically, “if you can sign up online, you must be able to cancel on the same website, in the same number of steps,” according to the FTC. Sellers must also “take ‘no’ for an answer” instead of continuing to pitch new offers when customers call to cancel a subscription.

“These companies are betting that customers will be too impatient, busy, or confused to jump through every hoop,” FTC chair Lina Khan wrote about the proposed rule.

To cement their standing on the subject, the FTC recently sued Amazon, accusing them of using deceptive practices that make cancelling subscriptions intentionally difficult for consumers, enrolling consumers in Prime without consent. An FTC chair pointed to « manipulative, coercive or deceptive user-interface designs know as ‘dark patterns’ to trick consumers into enrolling in automatically renewing Prime subscriptions. »

Reuters has talked about these so-called « dark patterns » before:

“Dark patterns are manipulative or deceptive practices built into user interfaces by developers that have the effect, intentionally or unintentionally, of obscuring, subverting, or impairing consumer autonomy, decision-making, or choice. Dark patterns are often carefully designed to alter decision-making by users or trick users into actions they did not intend to take.” 

Common examples of dark patterns include:

• Bait and switch: when information says one thing and a different action takes place upon clicking
• Hidden costs: the cost shown is different initially but increases as the buyer moves ahead with payment
• Triggering fear: when a user is asked not to opt-out of a subscription or feature as it can lead to negative consequences

Arguably, by making unsubscription more difficult than it needs to be, such as by adding in dark patterns or making it necessary to phone customer service in order to cancel, publishers are actually shooting themselves in the foot rather than increasing retention rates. Even if the reader may have been tempted by a reduced offer or month of subscription on-the-house, they’re so frustrated by the cancellation process that they’re put off the brand altogether. 

For example, an ex-subscriber to The Arizona Republic told Nieman Lab: 

“My wife wanted to go from twice a week to full seven days. Can’t do it by email with Republic. Told we have to call. Number to call seems to be regional phone center, constantly busy. Called Republic directly but were told to try to call earlier in the morning and maybe the call center wouldn’t be so busy. Always busy. Tried sending a letter to Republic, no answer. Tried sending letter to billing center, no answer. Finally went to bank and had auto withdrawal for monthly subscription canceled. Only response from Republic was that payment request was denied.”

Thomas Baekdal discussed this in his recent newsletter:

If you create an annoying canceling experience, like forcing people to ‘call us’ (which is now illegal in many places), you can turn an already bad situation much worse. For instance, if people are merely bored and you annoy them as they cancel, you push them into a « pissed off » mode instead.

You don’t want to ever do that. In fact, every single thing we know about churn indicates that the better the experience people have canceling, the more likely they are to either not churn if given the right incentive, or come back later.

So, you need to be incredibly helpful and respectful to mitigate churn.

To evidence this point, The Minneapolis Star Tribune found that allowing online cancellation, as it started doing in December 2021, hasn’t resulted in more cancellations. Instead, it may actually be helping the paper keep more of its subscribers who are on the fence. The Star Tribune’s “total online save rate” is 18.5%, while the “call center save rate” — where you have to talk to a person who tries to persuade you to stay — is only 8.8%.

Making it easier to leave convinced more people to stay in Norway

Norwegian Consumer Protection Authority determined that “it should be as easy to end a subscription as it was to subscribe in the first place”.

So, to adapt, the Norwegian newspaper Aftenposten made the unsubscription user journey shorter than ever – reducing the number of steps from 5 to 3, cutting all unnecessary steps and questions that might confuse subscribers and, contrary to all intuition, they made the cancel button available on subscribers’ « My page », meaning it is visible to them at all times in their account.

The subscriber cancellation flow:

When initiating their cancellation, subscribers are first met with information that they’re sharing their subscription with a friend and that their friend will also lose access if they cancel. (This only applies to those subscribers who share their subscriptions.) We know from previous analysis that this is a strong incentive to stay.

In the next step they will immediately be exposed to the holdback offer: “Can we convince you to stay with us for a little while longer?” The offer we’ve seen the best results with, in terms of volume, retention, and income is three months for the price of one. They are then presented with two choices: stay or continue unsubscribing.

If they continue, they are asked to choose between four reasons for their cancellation, amongst them an unspecified option to reduce the mental load, and afterwards, they get a confirmation.

This has actually had an impressively positive impact on their revenues: in six months, over 5,000 subscribers have been convinced to change their mind after starting the cancellation process, accounting for an estimated €320,000 (or US$344,000) in yearly revenues.

Tyler is the co-founder and CEO of beehiiv, previously being the 2nd employee at Morning Brew, building the infrastructure to grow and scale the company.

This article was originally published on Splice who work with the global media ecosystem to report on, teach, advise, transform, and fund media startups. Note that this interview was held in May 2022. Any out-of-date information has been removed or replaced & is in italics.

For someone starting a newsletter today, what are 2-3 things they should look for in a newsletter platform?

1. Scalability — will the platform grow with you (in costs, incentives, flexibility, data, etc.) ?
2. Complexity — Milk Road started on beehiiv with 0 subs in early 2022. 4 months later, they had over 130k.
   They didn’t need to build a world class website with video embeds, or overcomplicate things with tons of integrations and bells and whistles. Instead they plugged into our tech and let us handle web hosting, emails, data, deliverability, styling, referrals, etc. Their time is better spent writing good content.
   I think far too many people waste time spinning their wheels on things that don’t matter. I’d say, understand your competencies and maximize those, and find a platform that abstracts the rest.

> To add to your reading list: In conversation with 6am City – tech stack, audience strategy, and unit economics

What can newsletter platforms do to help distribute and market newsletters to other readers on the same platform? Is there a way to improve organic discovery?

So where beehiiv sits on this spectrum is purely on the Shopify / infrastructure side of things, not on the aggregator side. For example, when you shop at a shoe store hosted on Shopify they don’t tell you « you can also check out these 10 other shoe sites on Shopify » or « maybe you’ll like these shoes too! »

beehiiv is similar in that we provide the infrastructure and tools for you to spin up a top-of-the-line newsletter. Best-in-class deliverability, data analytics, subscriber attribution, etc. We also have a fully-integrated referral program to help you scale and incentivize readers to share and help you grow.

That being said, we’re not an aggregator. We’re not a destination for casual readers to find 15 different crypto newsletters. I actually think that part is solely on the creator/newsletter. I think you get in trouble as an aggregator suggesting certain types of content over others.we’re building tools for newsletters to connect with other newsletters and cross promote each other in some capacity, but that’ll come from the newsletters themselves, not beehiiv as a platform.

What advice do you have around pricing a newsletter subscription?

1. What’s a standard subscription cost for media / information / entertainment going for? In the US most subscriptions from music streaming to video range in the $8 – $15 range, so as a comp, that’s a decent place to start
2. How competitive is the space? For example, are you providing political insight that literally no one else is covering? If so, you could probably warrant a much higher price for that information because you’re a 1/1 (or at least 1/few). Alternatively if you’re just providing tech analysis, there’s a dime a dozen. Ben Thompson does a tremendous job, sending 3-4x a week, and charges $11 /mo, I believe. What’s the value and insight you’re providing relative to that?
3. Who’s paying for it? Is it consumer facing where individuals are pulling out their cards to pay, or is it b2b and for work? If the latter, if your newsletter provides value in a work function, it’s possible the employer will pay for it, so you can charge a lot more.
4. What’s the end game? Is subscription revenue your only revenue, or are you also selling ads? Are you selling courses? Are you cross promoting into different products? If you still want scale, but think you can warrant subscription revenue, maybe you intentionally price lower because you want to cross-sell other things.

How reliable are open rates given Apple’s privacy restrictions?

Simon Owens gives a more updated response here. In short, « Apple has declared war on email newsletters ».

« Apple recently announced new privacy updates to its iOS Mail app. Here’s how Digiday describes the changes:

With iOS 17 — which expands to public beta in July before the official launch in September — Apple will start automatically removing link trackers from URLs sent via Message and Mail as well as from links in Safari Private Browsing …

… Some marketers say removing URL parameters could make campaign analytics less reliable. Some of the “unintended impact” changes could include URL trackers related to ad measurement, embedded media, social widgets, fraud prevention, bot detection, audience measurement and funding websites that rely on targeted or personalized ads.

Apple is dead set on making it as difficult as possible for publishers and brands to measure the effectiveness of their newsletter strategies.

…Open rates and link tracking allow publishers to optimize their newsletters, and the latter also makes it easier for newsletter sponsors to track the effectiveness of their campaigns. By eliminating these features, Apple just seems to be giving more power to closed platform ecosystems that can track user behavior from end to end — platforms like Facebook, Instagram, YouTube, and TikTok.

Apple is basically hurting the decentralized web at a time when publishers are trying to become less dependent on the big tech platforms. I’m not sure consumers will be better off as a result. »

What should the key metrics be to track the success of a newsletter? Is it just open rates?

Depends on the type of newsletter. If you’re selling premium subscriptions, it might be the number of people who upgrade from free to paid, and the churn rate of your paid (you want to obviously minimize churn).

If it’s a free newsletter, then unique opens is probably important for you to monitor. Factor in unsubscribe rate, and lifetime value of the reader as well.

Also, if you’re selling ads, you’d probably prefer people who click and engage on ads more, because it makes your newsletter more attractive to advertisers (because they’re seeing success).

That’s a whole jumble of answers but truth is there aren’t a ton of metrics in email to start with, so know what moves the needle for you and optimize for that.

Have you seen any recent examples of unique/creative rewards for referrals beside the usual branded merchandise? Who does this well?

Going back to Milk Road as an example: they only have a single reward at 1 referral (an ebook) and they have over 10k referrals. They made it super simple, attainable, and it’s a slam dunk. Matthew Berry Fantasy Life is another fantasy football newsletter in the early stages of spinning up a referral program. They are offering unique fantasy insights and tips, and opportunity to speak with Matthew Berry.

What’s important is understanding your audience and aligning with what they want. It’s probably not another t-shirt or mug.

How would you go about growing a local newsletter email list on a tight budget?

• Great sharable content
• Cross-promote with other newsletters
• Experiment with incentives and rewards
• SEO
• Any sort of collabs in the community
• Partner with local businesses, if possible

> A 6-step plan for local news email newsletters to go from content strategy to growth

How does beehiiv work in segregating and teasing content for paid and non-paid subscribers?

Similarly to Substack, you can have a premium-only post. Alternatively, you can send a newsletter to both free AND premium, but drop a teaser line in the content so anything below the line is only accessible to paying readers, otherwise free readers are hit with an upgrade nudge.

In terms of engaging an audience, what are the challenges when the scale is so big? How do you keep the intimate flow of the conversation that characterizes a newsletter when the audience is so massive? Is that even possible?

One of hardest problems for a ton of businesses is scaling « exclusivity » or « intimacy ». I’m not trying to intentionally tie everything back to beehiiv, but this is something we’ve thought a lot about.

It’s not a perfect solution but it’s something: 1-click polls, essentially making it super easy to collect info and data on your readers.

It could be something as silly as « chicken sandwich vs burger? », but being able to then take the results and share them in the next newsletter is, I think, a cool mechanism for that independent reader reading in a silo to feel somewhat connected with other strangers reading the same thing.

The comment section is another, although I have mixed feelings about it. I think most online comment sections are mostly spam and not productive, but there are instances of great comment sections that I do think drive engagement.

How is the transition for paid newsletters from Substack to beehiiv? Do you have any mechanisms for potential subscribers to discover newsletters?

Something we prioritized early on was that you can migrate over from Substack in just 3 steps (~5 minutes)

1. Import CSV of subscribers (exportable from Substack)
2. Drop URL of Substack and we’ll pull over all your content
   • alternatively, if some of your content lives behind a paywall, you can import a zip file of your content (exportable from Substack)
3. Migrate premium subs (we have a mechanism to do this and our CTO will even white-glove and do it with you)

• Readers don’t get notified of any change
• Their billing cycle and payment method remains the same
• No disruption to existing subs

This article was originally published on Splice, who help to build viable media startups in Asia, working with the global media ecosystem to report on, teach, advise, transform, and fund media startups.

This is partly because of journalism’s roots as a purveyor of necessary information and the « build it and they will come » mindset – and partly because it’s a marketing discipline, not a journalistic one. Which is why some journalists and editors are prone to either ignore or forget its importance.

And even though product thinking and understanding of user needs are playing an increasingly bigger part in the industry (which is why this newsletter exists), old habits still die hard.

This means that old truths and assumptions still thrive in some areas. How to market and sell your content and subscription product (which is on track to become the number one revenue generator for many publishers) is one of these areas.

Fear of Missing Out

For instance, which do you think works best? Telling people they are missing out by not buying a subscription or that they are joining in and staying in touch with what’s going on and important by buying a subscription?

Usually the « fear of missing out » is a great driver and one that always works the best, some would probably argue. But is that true?

A study based on 11 experiments conducted in collaboration with three U.S. newsrooms tries to shed some light on exactly this: « The Effectiveness of Gain and Loss Frames in News Subscription Appeals ».

In short, « gain » is when you tell people what they’ll get if they subscribe, and « loss » is when you tell them, what they are missing out on if they don’t subscribe.

It’s not only publishers who have overlooked this aspect, though; « to date, research has not examined how digital subscription appeals affect news subscription intentions, » researchers Yujin Kim, Jessica Colliera and Natalie Jomini Stroud (all from The University of Texas at Austin) write in the paper.

(Stroud is a professor at the Department of Communication Studies and the Director of the Center for Media Engagement at The University of Texas at Austin, where Colliera is a Research Associate and Kim is a doctoral student.)

Yujin Kim tells me in an email:

So, what did they find out – which works best? Well, that depends.

Facebook vs. email

The experiments were conducted using email and advertising on Facebook, and the gain and loss framed messages were accompanied by control messages with text borrowed from standard advertisements previously used by the newsrooms.

On Facebook there was little to no difference. When using email, though, gain framed messages had a bigger impact (my emphasis):

« For the studies using email appeals, however, gain messages yielded 70% more clicks to subscription pages than did loss messages […] however, the difference was only significant relative to loss messages and not relative to the control message, » the researchers write in the paper.

This means that « stay in touch » appeals work better than « don’t miss out » in a direct email. Standard neutral messages work just as well as gain framed messages, though, this study indicates.

These results have previously been published in the ‘Subscription Messages’ report from Center for Media Engagement at The University of Texas at Austin.

So, why the difference between Facebook and email? And why shouldn’t you necessarily play on the « fear of missing out » when trying to sell your subscription product?

Prospect theory

The answers may lie in prospect theory, which was applied by the researchers in order to test the effect of the various messages.

The theory was developed by Daniel Kahneman and Amos Tversky in 1979 and was cited in the decision when Kahneman was awarded the Nobel Prize in Economics in 2002.

Prospect theory describes how we tend to make decision based on the level of risk and certainty involved. Whether gain or loss is more effective varies by topic, but generally the theory proposes that gain framed messages are better when there is little risk that an action will fail to produce the desired outcome.

In the paper, the reseachers give us an example:

« As O’Keefe and Jensen (2007) explain, dental hygiene is just such a case. There is little risk, they propose, that the advocated behavior (e.g. flossing) will not produce the outcome of healthy teeth and gums. The alternative – not flossing – is riskier. Gum disease is possible, but not guaranteed, if one does not floss. Under these circumstances, gain frames perform better. »

This means, in relation to news and publishers, that if the user feels pretty confident that what they are trying to achieve (by buying a subscription) comes with little risk, then gain framed messages will work better.

As this is the case in the study, it might give us an interesting glimpse of why people are paying for subscription products from publishers (my emphasis):

« Prospect theory suggests that variation in which frame performs better is a product of the anticipated outcome of enacting a behaviour and the certainty with which the anticipated outcome will occur […] In more certain situations, gain frames perform better, whereas in riskier situations, loss frames may be the better choice.

In applying the theory to the news subscription context, we must understand what outcomes people anticipate from subscribing and whether they believe that these outcomes are likely.« 

That makes sense. So how do you find out what people are trying to achieve by subscribing? This requires qualitative analysis, meaning that tools like Google Analytics won’t help you.

Jobs To Be Done

Understanding what outcomes people anticipate from subscribing and what they are hoping to achieve is usually addressed by using a framework such as Jobs To Be Done.

When using the framework you are trying to find and understand the user’s « underlying need » and the context in which the user is trying to get his or her jobs done.

A classic example in Jobs To Be Done is reading on the train. Today, many of us read on our phones, while 30 years ago we probably read books and newspapers instead. This is has to do with some of the underlying Jobs To Be Done: Passing the time.

Some might pass the time by listening instead of reading. They can listen to music (on the phone, perhaps Spotify, vs. walkman or discman earlier) or a podcast. And this is one of the core reasons why podcasts are popular with commuters; they address a Job To Be Done.

The framework focuses less on target audiences and demographics and more on the actual jobs and needs people are trying to find solutions to. In an MIT Sloan Management Review article from 2007, titled ‘Finding the Right Job For Your Product’, Clayton M. Christensen and others write:

Customers’ purchase decisions don’t necessarily conform to those of the ‘average’ customer in their demographic; nor do they confine the search for solutions within a product category. Rather, customers just find themselves needing to get things done. When customers find that they need to get a job done, they ‘hire’ products or services to do the job. This means that marketers need to understand the jobs that arise in customers’ lives for which their products might be hired. »

This is what the University of Texas at Austin researchers are addressing, and the combination of Prospect Theory and the Jobs To Be Done framework is a very powerful hybrid when analyzing and understanding user behavior and goals.

Forget key and target demographics and all the pretty slides for a moment and focus on what people are trying to get done – and why they might hire your product as a solution.

There are many ways to work with Jobs To Be Done. One method is to use the framework in tandem with the Value Proposition Canvas by Strategyzer (a model built on the Business Model Canvas), which includes « Customer Jobs » when trying to design a value proposition based on knowledge and insight about the users.

As the paper from The University of Texas at Austin hints, a great, well-thought subscription product should be based on understanding of these factors and drivers.

No matter how you choose to work with the framework, finding people’s Jobs To Be Done requires conversations, interviews and observations. Get out there and talk to the people you are trying to sell your subscription product to.

Relative certainty

That said, prospect theory can give you a sense of why people subscribe to your product. Using this, you can build assumptions and hypotheses on why people buy your product, which can be tested in both qualitative and quantitative studies.

Considering the results of the experiments, the researchers write (again, my emphasis):

Subscription decisions seem to represent the relatively certain and goal-oriented situations where gain frames outperform loss frames, as prospect theory suggests. […]

Not only does this research offer guidance on what appeals to use, it sheds light into why people are motivated to subscribe to news: the superior performance of gain frames suggests that subscription decisions are motivated by feeling relatively certain about gaining beneficial information from doing so. »

This is very interesting – and great news for publishers and the news industry in general. According to this study, people see little risk when trying to achieve their goals by subscribing to news – as they feel pretty confident that they will attain those goals with a subscription.

At this point the questions are plenty:

• What are they trying to achieve? Why?
• Do they indeed feel certain? If yes, why? If no, why not?
• How are they currently solving this problem? (Always a great question to ask)

…and so on.

(If anything, it’s a reminder that we all need to understand behavioural economics and the theories involved. Even if you don’t work in the news or publishing industry, you could probably increase your chances of success by applying prospect theory.)

A Word of Caution

They are, however, some factors which should be taken into account before applying gain framed messages to everyone everywhere.

The 11 experiments were conducted with three U.S. newsrooms. Out of those, just three experiments were done using direct mail. The remaining eight were done as sponsored posts on Facebook.

This means – as is usually the case – that more testing is needed. Let this be a call to all newsrooms and marketers out there: Test gain and loss framed messages compared with control messages.

Even if you perhaps choose not to learn from the results in this study, at least choose to learn from the method. That is also Yujin Kim’s suggestion to publishers and others.

« Some are already doing A/B testing on language appeals but my suggestion is that doing rigorous experiments (even though it’s on a small scale) with different messages/images or in a different medium may help them understand about their audience better, » she explains.

Conversion tests needed

And remember: The tests should involve more than just wording in emails and on other channels. As the researchers and newsrooms discovered, few actual subscriptions resulted from the experiments which focused on click through rate as its success parameter.

This was one of their biggest surprises in the study, Yujin Kim explains to me. Therefore she suggest that publishers think about how to use each channel most effectively.

« We think using different approaches is desirable depending on the goal of the newsroom’s campaign. For example, when it comes to new audience exposure, using images on Facebook can be effective, » she explains with a reference to a paper she has previously co-authored [see link at the bottom of this article].

« If they have already built some engagement with their audience (e.g., previous subscribers) focusing on benefits from subscribing to journalism is more convincing in emails. »

This rate of conversion need not be due to anything in the experiments, as the actual conversion from user to paying subscriber happens not in an email but on a web page.

In the paper the researchers write:

« There are many possibilities for this pattern. For instance, it is possible that the link included in the messages took people to a web page that was poorly designed, so that even those most interested in subscribing were deterred. It is also conceivable that none of the messages used in this study were particularly appealing. And another, more troubling, explanation could be that advertising for subscriptions via email and Facebook is not a promising strategy. »

The lack of uptick in conversions underlines that if you decide to test framed messages in your communication, you need to make sure to test the landing pages as well.

There are many factors to take into account, both in terms of messaging, wording and design and structure of the page where the actual purchase and checkout is taking place. And the user’s Jobs To Be Done, of course.

With that many variables in the equation, publishers and others should consider rigorous testing and learning through assumptions and validations.

What’s next?

The researchers propose some interesting questions to look at, though they aren’t currently looking into them. The most interesting one is perhaps looking at gains and losses in a wider, societal context – and not just a personal one:

« Recent research from the Knight Foundation suggests that statements about societal losses may increase subscription intentions. In their research, participants who were made aware of the financial crisis facing local news were more likely to donate to local news. »

Let’s hope that someone will follow up on this.

The researchers also suggest looking more into the difference between email and Facebook.

« For example, gain-framed messages may be more effective on email than on Facebook because Facebook yields a more negative mood. Studies have revealed a negative association between Facebook usage and users’ mood […] Although email can also produce negative moods […], other studies suggest that email can improve relationships and help people stay in touch. »

There is no doubt that we are just getting started looking into these aspects of news and the publishing industry. And with prospect theory and the Jobs To Be Done framework, success becomes more likely.