Will Mojeek Offer a Premium Service to Searchers?

This is a long post, so to summarise what I’m asking: will users be able to opt out of seeing ads on Mojeek by paying for a premium service?

Some important context that makes a strong case for why ads on Mojeek are a good idea: Ads on Mojeek

As well, a post about Mojeek’s business model: About Mojeek; Business Model, Surveillance, and Privacy | Mojeek Blog

This hasn’t been implemented in Mojeek yet, so I suppose I’m pre-empting it.

As a user, I see my search engine as a tool. I don’t want to see ads even if they are relevant. From years of using Google, I’ve long trained myself to completely ignore the first 3 results; it just means that there are 3 less results available on the page and I have to scroll further every time.

Even if I was interested in the ad, I would never click it—I’d type the address in manually because every time I’ve clicked an ad link, I’ve regretted it immediately. It takes me through 2 or 3 redirection hops and adds 50 unnecessary characters to the URL. I would happily pay $10 a month not to see ads.

From a small-medium business’s perspective, all of the methods for advertising on the internet are terrible. Whether it’s Facebook, Amazon, Google or Bing, they don’t care about you and you’ll likely have your ads taken down repeatedly, without warning, for a nonsensical reason, by an algorithm. They will often be reinstated, days or weeks later, having killed all of the momentum you’ve built up with your campaign—usually with a standard response and no apology. Amazon outright takes the money you hadn’t spent on advertising yet if your campaign was taken down.

Have you ever tried doing advertising for something medical-related? Not fun.

Not to mention that, as Mojeek says, “As an advertiser you would be feeding the beasts.” Mojeek Ads seems like a great alternative for a lot of businesses and is far more respectful of users.

Some Other Business Models For Search Engines

1. Kagi—Premium Search Only

Source: Why Pay For Search? Frequently Asked Questions (FAQ) | Kagi's Docs

Kagi strongly believes that advertising is the problem, so for that reason, they don’t implement any ads at all. However, all Kagi users must pay for a subscription (preliminary prices are $10/pm for a limited subscription, $20~30/pm for an unlimited subscription) to use the service. This means that Kagi doesn’t need to compromise between serving advertisers and users and can focus on building a better experience for users—if it doesn’t, users won’t pay for it.

They use Teclis, their own index, for indexing smaller non-commercial sites, and Tinygem for indexing news sites. They use Google, Bing, Wikipedia, and other sources for general results/instant answers. This also means that Kagi users are indirectly funding Bing and Google when Kagi sends a request through their APIs for every search. But they do deliver very good results, very quickly, and the interface should be entirely accessible without Javascript (settings are broken right now).

While I don’t like the idea of financially supporting and being reliant on Bing and Google (meaning you’ll get many of the same results, and much of the same undesired delisting: "youtube-dl.org" missing from search results - Kagi Feedback), I have to admit that I like this business model the most if users are willing to pay for it. But it means only paying members can access Kagi and they won’t generate any revenue from advertisers, meaning it is a model that will result in a decidedly more niche customerbase.

2. Brave Search—Ad-Supported & User-Supported

Brave Search/Tailcat is an independent index. Businesses can pay through Brave’s ad network for “privacy-preserving ads”. Users can opt out of viewing ads by paying $3 a month to become a premium member:

What is Brave Search? | Brave Brave Search is ad supported and, unlike other search providers, our ads adhere to the principle of privacy-first. […] Users who want to support the Brave Search mission, and get an ad-free experience, can upgrade to Brave Search Premium.

This seems like a good compromise, because most people won’t pay not to see ads, so advertisers still have access to a considerable database of users.

3. Google/Bing—Ad-supported, Search API Calls, and Directing Traffic to Other Products/Services

A business model many are familiar with. Searchers can’t opt out of seeing ads and there is no way for searchers to fund these search engines as a customer. Businesses are the only customers, and these “search engines have two ways to make money – get more users or sell more ads per user.” - Kagi. As a result, Google/Bing provide good enough results to keep searchers around, but don’t have any reason to respect searchers.

============================

So, I’m essentially asking if Mojeek has any plans to implement the same business model as Brave Search. As a matter of interest, I’ve also attached a poll to this post asking users if they would pay not to see ads on Mojeek.

Would you pay not to see ads on Mojeek?
  • No
  • Yes, for $3 a month or less
  • Yes, for $10 a month
0 voters

While I would personally not pay and be fine with non-targeted advertisements, but I am sure there are a lot of forks out there who are willing to pay for ad-free experience and it also enables yet another revenue stream for Mojeek. So it is always great to have both options, free and premium subscription model would be a great idea. Users with the free subscription should also be able to upgrade to premium anytime if needed. A great inspiration for this idea would be Neeva and brave search as you have mentioned.

But to make such a model a success, we need to make the premium subscription affordable as well, the less it costs, the more appealing it will be as it will just remove the advertisements for the users, so I’d say 2/3 dollars a month or 20 - 30 dollars a year would work out.

You might try Coil.

But to make such a model a success, we need to make the premium subscription affordable as well, the less it costs, the more appealing it will be as it will just remove the advertisements for the users, so I’d say 2/3 dollars a month or 20 - 30 dollars a year would work out.

One issue I alluded to when bringing up Brave Search’s model is that if every Mojeek user pays not to see ads, then advertisers have no one to show their ads to—and surely you would want to show your ads to the people who are willing to spend money? The less expensive the premium service is, the more users will pay for it.

In reality, however, I don’t think this is going to be much of an issue. I don’t expect most people will be willing to pay for a search engine in much the way that Internet Explorer forever killed the web browser market (and Netscape) in 1995—though not as drastic. Because search engines are a service rather than software, I believe there is still a niche, dedicated market willing to pay for them. So Brave Search’s $3 a month premium subscription should be a great compromise for their business model.

Thanks for the link! I actually read this post when it came out but had forgotten about it.

The Coil browser extension appears to be released under a free license: web-monetization-projects/packages/coil-extension at main · coilhq/web-monetization-projects · GitHub

The extension page on Mozilla Firefox says it is licensed under Mozilla Public License 2.0, but the LICENSE file for this repository specifies Apache 2.0 (while the extension itself appears to have no LICENSE at all), so there appears to be some confusion: Coil – Get this Extension for 🦊 Firefox (en-US)

Do you know if Mojeek doesn’t display ads if you are using Coil? I recall that Mojeek ads are in beta, so some users would be getting them (perhaps only in the UK), but I don’t think there has been any official statement on this. I wouldn’t be interested in using Coil to ‘donate’ to Mojeek in addition to being served ads.

The other thing is that Coil only works with cryptocurrency, which is why I initially ignored it, as I didn’t want to go through KYC for any exchange.

Honestly speaking, in today’s world, especially as we grow, there will always be people who don’t care to see 1 or 2 non-targeted advertisements and really won’t pay for a premium subscription model, even if it’s dirt cheap like 2 or 3 dollars a month, because to some, they don’t “care” about these ads, so the majority today won’t pay for it in the first place. So that is why make it affordable, because there are always niche people who are really concerned about ads affecting their search results and want “organic results”. I think the majority though, won’t care to spend 3 dollars a month, atleast maybe average users. So at the end of the day, premium subscription model will be yet another revenue stream that can bring good profit, but no matter how affordable it could be, there will always be people who don’t care to spend a dime to remove ads either.

If you go to brave search discord, the majority there think having premium subscription option is a good thing, but many still use the basic free version, so I think both models can survive at the same time.

I guess this will depend on how ads end up being implemented.

The help topic below makes it sound like they accept credit cards.

https://help.coil.com/docs/general-info/intro-to-coil/index.html#does-coil-have-an-ico-or-token


@Archit I think the people who come to Mojeek are making a conscious choice to remove targeted advertising from their lives. Mojeek currently offers contextual advertising. And web monetization is another way to opt-out. I want to support Mojeek without the negative consequences of the current big tech ecosystem.

1 Like

Paying a subscription fee means I have to register, which sort of defeats the whole purpose of a private search engine.

Non-tracking ads don’t bother me unless there are too many or they are too annoying and then I would go elsewhere for searching.

Personally, I’m trying to keep registering at sites and also keeping the number of subscription fees to a minimum. On subscription fees alone: my car, streaming services, news sites, software apps, all want a monthly subscription fee which would put my credit card and identity information all over the place. There are just too many places with their hand out for a subscription and it adds up plus I would have to keep track of all of them.

3 Likes

Now that is a good point, didn’t think of the registration point

Paying a subscription fee means I have to register, which sort of defeats the whole purpose of a private search engine.

I have to disagree on this one. Mojeek has every opportunity to track you regardless of whether you pay for the service or not, just based on your IP address. Ultimately, this relationship is based on trust. You could make the argument that being logged into an account makes it easier to tie different IP addresses to your identity, but if you’re switching between identities using a VPN, you should be using a different browser for each identity (you don’t need to be logged into Mojeek on the other one).

Mojeek may not require you to “register”, as such; they may instead provide you with a 16-character token on the end of the URL that is valid for 1000 ad-free searches. You might be able to pay with cryptocurrency or by cash order. I’m using Mullvad for inspiration here: Mullvad - Wikipedia


Now, on subscriptions, I very much agree, but this is one of the rare things I would happily pay a subscription fee for. I’m part of that niche market that would happily pay to sustain a search service in return for a premium experience. Instead of a subscription, you could pay for a number of ad-free searches, which could be a one-time fee.

Coil uses confusing language that is not entirely clear, making constant reference to a “ledger”, “wallet”, and “digital currencies” which is terminology commonly used for cryptocurrency. Here, they say (Open Letter):

your wallet provider must support Interledger. Uphold and GateHub are the only providers using Interledger at this time.

Here, they make it clear that they use USD but use cryptowallets as the delivery mechanism: Open Letter

Coil pays out in USD. What happens after that depends on your digital wallet.

Can I use my bank account to receive payments from Coil?
No. Unfortunately, banks don’t support the Interledger technology that Coil uses to send payments.

You would still need to go through KYC for Uphold/GateHub, so Coil doesn’t accept credit cards through Paypal/Stripe. Having actually gone through KYC for two cryptocurrency exchanges already, it’s not an experience I really want to repeat…

1 Like

On Coil and its relation to crypto wallets etc. it essentially works like this:

The Sender

Will pay a subscription to a service such as Coil, Web Monetization is proposed as a W3C Standard, so the idea is that Coil would eventually be one of many providers of streaming payments.

The subscription is paid via a bank card, which obviously has privacy implications, but I am unaware of any need or possibility for using crypto on the sender’s end.

The Medium

The browser extension you activate once you’ve subscribed allows anonymous streaming of payments through an intermediary (in this case Coil) using the Interledger Protocol.

The WM Spec might be of use here.

The question from the FAQ that concerns blockchain:

Does Interledger Utilize Blockchain?

The Interledger Protocol is not based on blockchain, but it uses some key concepts from blockchain technology, such as a decentralized design and cryptography-based security.

According to the Interledger white paper, one of the most important objectives of the Protocol has always been to create the necessary conditions so that “any ledger can integrate to this Protocol simply by enabling escrowed transfers. Unlike previous approaches, the Interledger Protocol does not rely on any global coordinating system or ledger for processing payments – centralized or decentralized.”

The Receiver

They will have these payments streamed into their wallet, for which they do have to go through KYC and have it with an approved provider. When the money comes in it goes in as fiat. I believe you can ask for it to come in some other way, but for us it is just USD - this was the default and I have not changed it.

This all being said, looking at the language, they could be clearer.

2 Likes

Thanks for that. The explanation you linked/provided is a lot clearer than what Coil presents you with (and I also find their site quite confusing to navigate). I see I actually misread part of what I was quoting. So I did end up signing up for Coil with my credit card (they use Stripe) but wasn’t particularly impressed by it.

Some of my impressions:

  • The Coil Extension and site don’t tell you how much of your $5USD has been spent, and on which sites. I looked at the documentation, but there doesn’t seem to be anything related to this.
  • There appears to be no way to limit Coil to only pay for sites that you want to pay for (aside from not visiting them). I don’t want to pay for every site I visit—in fact, I don’t want to pay for most sites.
  • Coil will automatically deduct $5USD the moment you hand them your credit card details and offers no refunds.
  • The sign-up process is very painless. I was able to sign up, install the extension, and get it working in 3 minutes.

Then again, it’s still in beta, so these things can still be improved. Judging from what I’ve read of Coil’s site, I don’t think they want to allow the user to choose which sites they pay for, however:

With services like Patreon, you select which creators to support, then pay each creator separately, depending on the membership plans they offer. Coil streams payments in real time to any web monetized sites you visit.

I don’t think the Web Monetisation spec is a bad idea, but I think Coil’s implementation needs improving. However, I don’t think there’s much issue with paying via debit/credit card. The $5 goes straight to Coil’s Interledger wallet to be managed, which means that your bank (and whoever else follows the paper trail) doesn’t know which sites you’re visiting; only that you paid to use Coil. Coil, of course, knows which monetised sites you’re visiting and when.

1 Like

Yeah I think you are quite rightly pointing out how the explanation of something could be way more clear and concise.

From talking to them, there is no upper limit. It’s a subscription that covers you for all browsing across the course of a month. I guess they’ve done calculations on their end that say there will always be a bit of money left over even if the micropayment streaming is left on 24/7 as some people have tried before with streaming services.

Obviously depending upon the tech playing the way they say it does (something I am very much not in a place to verify) they have a system for preventing that. The CTO’s article is better than these FAQs; as they say, the data available to them with their combination of tools is that:

Coil learns that an unknown user is visiting some site which uses ExampleWallet as its wallet.

It’s all very interesting, but I haven’t really seen much change since we first came across it. Hopefully there will be more to come, with no more KYC, I feel your pain there.

1 Like

From talking to them, there is no upper limit. It’s a subscription that covers you for all browsing across the course of a month. I guess they’ve done calculations on their end that say there will always be a bit of money left over even if the micropayment streaming is left on 24/7 as some people have tried before with streaming services.

I think this is fine, but I was really expecting a breakdown similar to:

You have $3.09 Left.

This Month, These Sites Received Money From You:

  • Mojeek $0.93
  • Hackernoon $0.42
  • Imgur $0.02

But they give you absolutely nothing, not even telling you which sites have received money from your visits:

Confusingly, I suppose my membership could be inactive, but I could still have $3.09 available and not know. They’re just very sparse on details for the visitor. It seems like they give receivers a lot more detail.

Obviously depending upon the tech playing the way they say it does (something I am very much not in a place to verify) they have a system for preventing that. The CTO’s article is better than these FAQs

If that’s the way it works in practice, then that’s great. One less thing to worry about.

I might warm up to the idea of Web Monetisation (and Coil) as it develops further, but I don’t know if I’ll want to keep another extension installed just for this. I can see it as being something Mozilla might integrate into Firefox Stable eventually: Innovating on Web Monetization: Coil and Firefox Reality - Mozilla Hacks - the Web developer blog

And, yes…if I never have to do KYC again, I’ll be happy.

There seems to be some confusion.

An account balance like $3.09 or $0.93 requires identification.

Coil focuses on privacy by blinding themselves to both the sender and receiver. The CTO’s article Josh mentioned helps explain this.

When the extension is started, Coil issues it several signatures that say,
“I’m a Coil subscriber.”

When the extension loads a site, it redeems the blinded “I’m a Coil
subscriber” signatures. Coil verifies the signatures and starts sending Web
Monetization micropayments to the site.

Coil doesn’t learn which user triggered the Web Monetization micropayments,
because you’re using blind signatures.

So, your options are “subscribe to Coil” or “not subscribe to Coil”. Actions like checking your balance or pro rata refunds are not possible because of the focus on privacy.

Mike

1 Like

I have to admit to skimming the article Josh posted at the time (though I plan to read it properly later). Thank you for clarifying why there’s no account balance and no refunds. The compromise for privacy is inconvenient but understandable.

However, the extension still tracks your browsing history, so at the very least it should be able to tell you which sites you’ve paid this month. And since it’s measuring the time you spend client-side, it should be able to tell you how much time you’ve spent, or even estimate the amount of money paid. To clarify, $0.93 wouldn’t be a balance, but it would be the value of the amount of signatures you’ve redeemed, which should be easily trackable by the extension in theory. However, even if the value is not easy to determine, it should be simple to determine how much time was spent on the site.

And there shouldn’t be any technical limitation regarding being able to control the sites you pay for with the extension, unless I’m misunderstanding something.

Honestly the killer feature of account-based search engines is custom ranking. Kagi and Neeva let users block, demote, and promote certain domains. Demoting w3schools, dev.to, geeksforgeeks, and Quora while promoting MDN, w3c, WHATWG, and official documentation for programming langs has significantly improved my result quality. I still often use Mojeek (and others) for general search but I reach to Kagi for programming and non-commercial discovery.

1 Like

Premium Service and Business Model

This long post is a response to several points made above. Thanks for sharing your thoughts, information and your views; this is always interesting and helpful.

My initial task when joining Mojeek as CEO in July 2020 was to decide what our business model would be, and what actions we should prioritise to make progress as a sustainable business. With a full-search-stack, independent IP and our own infrastructure we had, and still have, numerous paths available. We explored and considered many options, to a greater or lesser some extent. Here I will summarise our decisions, actions since, and something about where we are heading.

Whilst building and expanding our free web search services, we decided to focus on two things, initially:

  1. Contextual, no-tracking, search ads

  2. Web search API customers

whilst leaving ourselves to pursue a third option, in the medium term:

  1. Subscription service option(s).

We also decided to experiment with Coil as a micropayments service, as linked to by @mike, but more out of curiosity and a wish to support their goals. Since the objective of Mojeek is to get you quickly to search results, a time-based micropayments service is counter to that. @Josh has hopefully clarified how Coil works. The fact that it is an open protocol, and not a crypto based operation were two big reasons to try Coil rather than explore other options. Using Coil brings us a tiny amount of income but that is not our motivation here.

1. Search Advertising

We have made good progress on our contextual search ads product, with paying Beta customers who have also helped us develop the product. We built our own ads technology, so that we are not dependent on Google or Microsoft ad networks, nor indeed anyone else. Since ads has been discussed here (by @gnome, @archit, @mike, @brad) I will make some additional comments about this.

As we said in the post referenced [Ads on Mojeek] (Ads on Mojeek), Ads “can be a useful economic signal for a buyer and a way for upstarts and smaller brands to get attention and compete.” We believed that then, and contact by people and SMEs since we launched our Beta programme, has confirmed that hypothesis.

With regard to subscription/premium services: we believe these can be, for one thing, discriminatory. We believe that information wants to be free, and that wherever possible and important, information should be free available and accessible by individuals. That’s an ideal, and obviously does not necessarily relate to B2B services. Still Mojeek as a B2C search service should always provide some level of free service which is accessible by anyone. If it takes contextual ads to support that, then so be it, within our guiding principles.

If we return to a society where only a (richer) subsection of people can search for information on the web, then we are heading towards a new dark ages. There was a time when books where accessible only to the rich, elites and religious establishments. We are very conscious of a similar dangerous now digital trend towards a dark cyber age. We assume that we are fighting, and with you, against such forces and trends.

It’s fine @gnome that some people never click on Ads. I’m almost one of them myself. I say almost as I must admit that I will have clicked on a search ad at some point in may years, though probably more often than not accidentally! As we have said and will keep saying “Ads are not the fundamental problem. The fundamental problem is tracking.” Tracking is the main tool of the digital oligopoly and surveillance capitalism; as best understood and expressed by Cory Doctorow.

We couldn’t be an adtech company even if we wanted to be one. To be an adtech company we would need to switch to an operation where we focus on harvesting and using data about you, directly or indirectly through fingerprinting and data trading. That goes totally against our principles, our proposition, and founding principles. Values that sustain us and keep us working on Mojeek every day.

2. API

We have significant revenues now from our search API. In some cases we have been able to succeed by providing capability and terms not offered by Google or Microsoft.

To pursue the subscription options we would need to add value, over and above that which is provided by our free service. Now if you are taking the short road to provide a premium search service, then you can get there quickly if you are willing to pay Google or Microsoft for their APIs. Kagi, Neeva and others as mentioned by @seirdy have done that.

Of course, Mojeek API customers can also take a shorter road of building a premium service by using the Mojeek API, and that would help us and them.

In our case, we are taking the long road and building from the ground up. We also do not wish to support surveillance capitalism companies that compete in search. So building a premium service using Bing or Google, or even Yandex or Baidu for that matter, is not us.

3. Subscription Service

On our chosen path of independence, we have been building the ability to explore premium/subscription services. We will have specific news to share on that very soon; a little taster, if you have not seen before, is here. It is a new product that we restarted work on last year. Version 1 is now nearing completion. The roots of this go back to February 2006 when Mojeek “Personal Search” was announced. More details were reported on here in September 2006. This project was put aside to concentrate on building the Mojeek general web search engine that you use, and we run today. The new product is something that we are intending to bring to this community first, ahead of any general release and after internal testing.

5 Likes

Thank you @Colin for this post. Well said, all of it. Here’s my quick thoughts.

Search Ads:

Ads without tracking are much needed on the web. We can talk about the evils of tracking ads networks all day, but nobody can move away from them until non-tracking ad alternatives become available. So thank you for building this. Now we all need to attract more eyeballs to Mojeek.

a way for upstarts and smaller brands to get attention and compete.

Again, I think this is important. The big engines really only want big spenders. They’ve forgotten or don’t want to be bothered by the little person.

Aside: Back in the day it would take at least a month for Google and or ODP to list a new site. So when I launched a new website I’d spend a whole US$20 at Goto .com (a bidding search engine) using 1 or 2 cent keyword bids (many long tail) just to drive a little traffic to test things. I was the smallest of the small context ad buyer but Goto .com was not too big to accept my tiny $20 bill. /aside

API:

I’ve said it before I’ll say it again: if you combine Bing, Mojeek, Gigablast and Yandex together you could have a good metasearch engine like Ixquick or Clusty. That was before Brave search and Kagi were around they may have potential for that list. I’m not sure how deep their crawlers actually index but I’ll leave that to @Seirdy to suss out. Someday I expect Mojeek to supplant the need for Bing from that hypothetical meta search list.

And I do hope that some current meta search services will quit being just Bing retreads and use some of the smaller engines. Especially if that gives Mojeek both income and exposure.

Subscription Service

I like this. It’s got to be more than just getting rid of ads for a subscription to be worthwhile. (I loved Rollyo back in the day and created a couple of really good niche engines using it. Like everything Yahoo! ever invented they never went anywhere with that and it died.) I’m not sure Personal Search is for me but I can see it being really good for some people.

Misc.

I miss the old Pandia search engine reporting as cited in your Archive links.

Don’t ever IPO Mojeek. It was the kiss of death for every search engine/portal that ever did it. And it was the “seeds of their own destruction” for Google. Better to stay closely held.