The internet today is broken. In the race to more effectively deliver ads, we have missed out the larger goal of the internet, which is to create meaningful connections with other people and drive the production of knowledge. Much like how badly designed incentives brought the financial markets to its knees in the 2008–09 crisis, the incentives inherent in the ad-driven internet are slowly pushing us towards an increasingly complex system where only the middlemen benefits at the expense of companies, creators and users.
On average, there are more than 40 trackers on any news site that you visit today with over three quarters of them purely there to measure you preferences as a consumer. Imagine if for every newspaper article you read, there were 30 pairs of eyes peering over your shoulder, judging you and trying to categorize you into one of their demographic buckets. We would definitely take notice if this was the case in the physical world, but in the virtual world, the lack of any physical feedback makes us ignorant to the amount of personal data we are allowing people to collect. Even for those that do take notice, many dismiss it as just the trade-off for a ‘free internet’ where we trade our data for services. As such, to tackle this free lunch problem, it requires us to first take a step back and answer the following question:
How do we fund the internet of the future?
The current market: An exercise in futility
The increasing complexity within the ad-based marketplace has led to single ad units being bounced across multiple networks as the market tries to find the ‘best’ target for the ad unit. Along the way, the ad unit will incur additional market-maker fees as well as pick up additional data that is tagged to it resulting in increased energy costs with transferring this data across multiple databases.
For the consumer, this has resulted in longer page loading times as well as increased costs incurred through additional data being downloaded. Even more concerning is that at each stop that the ad takes, the consumer also pays indirectly through the selling of their personal data. Consequently, privacy is also a big issue especially with the public becoming more aware of it following the Cambridge Analytica scandal. It is hard to say that the experts didn’t see it coming as this is the natural outcome when the incentives push the industry to micro-targeting, especially when there is little oversight from governments which have always lagged technologically.
Related to the rise of consumer data as keys to profiting, platform delivery channels have also become more centralized as platforms such as Facebook and Google are increasingly getting the lion’s share of consumer data. This is a double-edged sword as this data enables services from these platforms to be more personalized but there is a real risk of stifling alternative views with this top-down power structure. This could be seen in the YouTube demonetization episode whereby companies started pulling out their ads as they didn’t want to be linked to controversial views. The ones that were hardest hit by this were the content creators resulting in less discourse around alternative media.
Aside from the risks of ad fraud and malvertisements, pay-per-impression incentives have also lead to what I would argue is an even bigger problem: click-bait. This not only frustrates users but in aggregate, huge amounts of resources are being wasted on generating content which have no value except to get clicks or in the worst case when it is in its most effective state, to incite controversy and anger between two groups (especially between political groups).
When did ads become creepy?
If you think about traditional ads in the past, these ads were distributed in the places where they had the highest visibility. This could be the front page of the local newspaper, the billboard beside a busy highway or even the commercial break between ‘The Sopranos’ show in the early 2000s. Targeted ads were limited then as the most they could know about the customers were that they like watching people getting massacred by a cigar-smoking, suit-wearing and Italian-sounding men. Even then, I imagined the range of products they could sell were likely very limited. This is unlike today where a simple google search of healthier food options result in an eternal bombardment of weight loss products.
The internet was the key that enabled companies to start forming a virtual identity for all its customers which was also greatly aided by the rise of social media profiles. This was even more so with the rise of personal smart devices since the launch of the iPhone. The wealth of data companies could have only dreamed of a few years prior were suddenly made available. The companies which could sort through the glut of data were best positioned to tackle the future as the oil that drove their profits was essentially free as consumers were slow to realize that they left a digital footprint.
It is easy to say in hindsight, but the likely outcome of an incentive structure based upon matching ads to target consumers, was one where a marketplace would spring up to harvest as much user data as possible. Consequently, price discovery as we know it in the traditional economy would always require a trusted third party due to technological limitations. This added on layers of complexity in the ad marketplace as companies started seeing profits from optimizing their own matching algorithms. As with any system, the more complex the system got, the more potential points of failure were introduced into the system.
What alternatives do we have?
- Pros: Users don’t incur a cost directly; Companies now have more data which can provide insights to how society functions, but these are hidden in their respective silos
- Cons: Users data becomes a commodity on the ads marketplace; User’s ownership of their own data becomes a grey area; Increased latency when loading pages; Data carrier charges for additional data loaded per page; Pay-per-impression leads to click-bait; Centralization of content platform delivery channels
Freemium (Content has a paywall with free users getting access to non-exclusive content)
- Pros: Companies are able to determine which content is publicly accessible; Through conversion rates, companies are able to continuously test price points
- Cons: Favors the bigger players due to their breadth of coverage; Content is limited to users who can pay the fee; Content is siloed
Freemium (Content has a limited paywall with free users getting access to a limited amount of content)
- Pros: Allows the user to choose what fits into their content limitations; Companies are able to see true on-site behaviour as there is no restricted content
- Cons: Favors the bigger players due to their breadth of coverage; Users who are unable to pay will have to consciously navigate given their content limits
Subscriptions (Users pay a fee to gain access to the content for a fixed period of time)
- Pros: Predictable flow of revenue for companies; Optimal viewing experience for users
- Cons: A pure subscription model is not sustainable as brand discovery is not possible; Subscriptions can work out to be very costly if the user only requires access to a single product/content
Tipping platforms (Users have free access to the content and are able to pay the content creators directly through tipping mechanisms)
- Pros: Users are able to support the creators they like directly; Less resources will go to the middlemen with creators benefiting from this improved efficiency
- Cons: Costs of doing micro-transactions make this less feasible although potential use of cryptocurrencies will reduce transaction costs significantly; Favors content which targets users who can afford to tip; At a higher risk of a ‘winner take all’ situation if algorithms are not optimized for content variety
- Pros: The content itself matches companies to the relevant users; Research has highlighted that local influencers are able to obtain a higher conversion rate as users place more trust in a familiar face
- Cons: Content will be curated by companies; Resources get concentrated to top influencers as number of followers replaces content quality as a key metric for companies
In-browser crypto mining
- Pros: The least intrusive as there is no identity tracking required; Creators are incentivised to create more engaging content as the amount of time a user spends on the site decides the amount of crypto mined
- Cons: As mining can be done on all smart devices, it will favour users with more efficient devices; Costs involved with energy and data transfer; Decreased device lifespan due to throttling
As can be seen from the above, many of the models have its own trade-offs which is why the ad-based model has stuck around as the incentives for change have not been large enough to force people to move towards another model. In the following section, we will take a look at some innovative solutions some companies have come up with to tackle this problem.
Medium has managed to get rid of ads completely while still allowing ideas shared on its platform to be freely accessible. The Medium model enables all content on the platform to be viewed by anyone but limits the number of premium articles for non-paying registered customers to 3 a month. For users who pay the $5 monthly subscription, they will have unlimited access to all premium articles plus a few extra perks such as audio-narration and offline reading.
One vital thing Medium has done in empowering the content creators is giving them the ability to decide which of their content is considered premium with creators only getting paid for premium content. In terms of how the creators get paid, the $5 subscription fee is distributed proportionally at the end of the month based upon the reading time and number of applause (Medium’s own voting mechanism) a user gives to all premium articles they have interacted with for that month.
The Medium model manages to balance the need for content to be freely accessible while also enabling creators to get paid. This being said, there are also a few issues with this model. Key amongst this is that a paywall still exists with articles on the platform being split between premium and public articles. Given that the only revenue stream which Medium has is through its subscribers, it is understandable why this needs to be done but it comes at the cost of ideas being siloed. As such, there is a fine balance to be had for displaying premium and public content to non-paying members. For those that don’t want to pay or who struggle to pay, they must consciously choose the premium articles they want to see.
Related to the monetisation of premium content, another challenge is that creators only get paid for content which they deem as premium. Putting aside the fact that it is very difficult to determine which of the articles will go viral, this model requires the the creator to decide on the tradeoffs between monetising their content or gaining more exposure by releasing it for free. Ideally, monetisation and exposure should go hand-in-hand as creators should be rewarded if the knowledge shared for them is of value to readers.
One major challenge for Medium as an idea sharing platform is that anonymous participation is not possible given that the payment channels powering it are only credit cards or PayPal, both of which will effectively link the account with a user’s identity. It is still possible to participate with an anonymous e-mail account but such users will not be able to give or receive any monetary payments. Given that Medium is a single platform, this also means that the platform is susceptible to censorship as seen in Medium being banned in China and previously also in countries such as Malaysia. Censorship resistance will be key to the future of the internet if ideas and discourse is to be encouraged.
Brave takes a rather innovative step towards providing users with more control over their data and how they choose to fund their favorite creators. Brave aims to monetize publisher content while protecting user privacy. To do this, they have rolled out the Brave browser and the accompanying Basic Attention Token (BAT).
The Brave browser acts as a gateway to the internet for users, much like Google Chrome, Mozilla Firefox or Microsoft Edge (Internet Explorer). For a user on the Brave browser, the experience of navigating the internet will still be largely the same albeit much faster, secure and private as the browser automatically blocks third-party ads and trackers.
The biggest innovation that Brave brings to the table is that users will not have to change their web-browsing behavior and are even incentivised to switch to the Brave browser, all of this while protecting user data and driving efficiencies in the ad marketplace. This is achieved via the Brave browser which privately monitors user intent at the browser level and encrypts this data to be stored on the device only. The browser tracks engagement via various algorithms which calculates the amount of time an ad on an active tab is in view. In-device machine learning will match relevant ads while companies will be able to buy ad space via a decentralized ad marketplace.
Additionally, these micro-transactions are made more efficient via the Basic Attention Token (BAT), an ERC-20 token. By buying an ad-space, advertisers deposit BAT into a locked state which only opens when a user views the ad thereby allowing money from that account to be distributed to Brave, the publisher and also the users’ wallets. As such, BAT is effectively a monetary valuation of a user’s attention.
Key to Brave’s success is that their revenue streams are not just limited to ads as BAT empowers creative content by enabling users to indirectly contribute towards their favorite creators. User’s of the platform have the option of setting aside a monthly contribution amount which is divided proportionally among the registered Brave publisher sites they engage with the most for that month. This evens out the playing field as content creators do not have to be locked into the pay-per-impression cycle.
A possible pain point in this design is that for content creators, they will have to register with Brave as a publisher hence Brave still has the power of verification and censorship. Moreover, as it stands now, the in-browser Brave wallet is currently only unidirectional with future updates for multi-directional transfers requiring know-your-customers procedures. This being said, the Brave infrastructure is open-sourced and Brave as a whole have been very transparent with their team and direction. Brave is planning to move towards a fully distributed micro-payment system in the future and this is exciting as creators will have access to more revenue streams.
Steemit is a social media platform where everyone gets paid for creating and curating content. Steemit flips the conventional value exchange system on its head by getting the community to pay creators for content which the community as a whole deems valuable. This is made possible through a relatively complex platform architecture which is powered at its core by STEEM tokens.
Without getting too much into the technical details, users are able to upvote or downvote content on the platform and the distribution of these votes will determine how the rewards pool is distributed to the creators and the promoters of content. The users influence on the distribution is directly proportional to the amount of STEEM tokens they have vested in the network. The rewards pool is not created out of thin air but rather is the result of users vesting their STEEM tokens in the network. At a simplified level, this rewards pool can thought of as a portion of the interest generated by the network once all interest payments have been paid to individual users who vested their tokens (for those familiar with blockchain, this is basically the block reward). As such, when a user upvotes a post, it is the community that pays the bill.
As groundbreaking as this innovation is, it also comes with some drawbacks. Chief among them is centralization risks as a users’ voting power is directly proportional to the mount of STEEM they have vested in the network. As such, larger holders of STEEM are able to disproportionately influence the public discourse as their votes will have a significant effect on what content is promoted or hidden from sight. There is a lot of discussion around this but it is a problem which is solvable through limits or power delegation. Steem themselves are trying to make distribution fairer via delegation of votes.
Additionally, Steemit faces the problem that this ecosystem is only accessible through a single platform. As such, it requires users to buy into the network although there is an option of signing up anonymously. Nonetheless, this will be a significant barrier to adoption for many users as not only is there resistance to change but the content itself is limited to the Steemit platform.
Lessons from Medium, Brave and Steemit
Key to all of the companies explored above (save for non-paying users in Medium) is that the users do not have to make a payment decision on a per-interaction basis. As such, the user is able to freely navigate the platform/web without having to worry about incurring additional costs. This is vital to the future of the internet as it is not necessarily the monetary costs which users will worry about but rather the uncertainty that accompanies deciding whether a site is worth paying for.
Moreover, through distributing resources via the various voting mechanisms, the incentive is for the community to continue creating engaging content which is valued by the community as a whole. This subjectivity in what constitutes value in the network will enable the ecosystem to continue growing.
Given what we have explored above, there are a few design principles which I believe will aid in creating a better internet:
- Save for content such as financial trading pieces where their value is derived from exclusivity, content on the internet will have to remain free to reduce decision anxiety around micro-payments. This will ensure that every one that has access to the internet will also have equal access to knowledge. Effective content and brand discovery also hinges on this idea.
- Given that digital assets are very easily copied, content immutability will also be key in enabling data to be freely accessible. This will ensure the original content creators are the ones being rewarded for their hard work. As such, the blockchain will likely play a big role in content distribution in the future.
- There will be a need for a voting mechanism that indicates what users value on the network. This will enable the ecosystem to continuously evolve in alignment with what the community as a whole deems valuable.
- Participation in this network will be based upon how much each individual user is willing to contribute to the communities which they participate in, be it through monetary (contribution pools, tipping or buying affiliate products) or non-monetary aspects (labor or personal data).
- To ensure a thriving ecosystem, the network will have to subsidize some of the costs in order for alternative views to be visible on the network.
- Ads will likely still have a part to play as the consumer and the companies still benefit from this relationship. However, the marketplace for ads will probably operate on the basis of minimal amount of data required. Additionally, this data will have to be anonimized and owned by the users themselves. Transparency over how this data is being used will also be key in gaining consumer trust.
- Value-added services will be important towards the sustainability of a platform but this must be optional and not come at the cost of platform usability. For example, paying for additional storage space or even stickers brings real value to those that want it without compromising usability for a non-paying user. Marketing of these services will have to be balanced with ethical design choices.
The principles above form only the starting point on how the internet of the future should be designed. I would love to hear your thoughts on this and other ways we can build a better internet that encourages meaningful interactions.