I’d argue for just considering technical criteria rather than how the company is set up or funded. Just add a warning banner for new companies/services.
1 Like
I don’t think there are any reasonable metrics which could be added to predict the future stability of a provider.
What should be added is an evaluation of how painful it would be to change if the provider becomes no longer recommended. This would probably be both an evaluation criteria and a recommendation/advice to make it easier to change if this becomes necessary.
I think if a recommended provider changes to no longer recommended it would be nice of PG to provide a “easiest way to switch” guidance as less experienced people may spend a lot of time finding the steps to take.
I also agree with the idea of having a list of no longer recommended providers/apps with reasons why. In fact a list of all providers/apps which are not recommended (never passed the test to get to recommended) and why would be useful.
Finally I don’t think VC funding is a reasonable criteria for rejection. I have worked for a startup where the VC was low pressure and the intention was for the company to pay out the VC as they grew, which the VC had agreed to. However having a recommendation would have helped bring customers in which would have helped grow the company in the first place.
I
3 Likes
I think a small warning banner for new companies / services with less of a track record is a sensible solution here.
I think a lot of people are missing the suggestion. Jonah isn’t saying that PG should never recommend products that are VC funded. He’s saying that companies that initially rely on VC have three only possible fates (being bought out, IPO, or buying out the investors), and until it’s clear which of these three is going to happen, it’s risky to recommend anything because they have to do one of these three things eventually. Proton was initially funded by VC, yeah, and it shouldn’t have been recommended at that point. It was recommended after that point. Similarly, perhaps PG shouldn’t be recommending products until the final business model is settled. A settled product that already has a sustainable business model receiving further VC funding isn’t what we’re talking about.
I disagree that PG should care about up and coming products, or that it’s a loss to the community if PG doesn’t recommend potentially great startups. I don’t think it’s PG’s job to advertise. Their business model is their problem.
I disagree completely with people suggesting that Skiff was an unpredictable one-off. Absolutely no it wasn’t, we all knew it was sketchy, we all knew they were pushy, and it’s not surprising at all that it was removed. It should have never been recommended because the business model could never have worked, because it wasn’t the final business model (again, VC, it had to change somehow).
I also disagree with the idea that, well, just remove them from PG when they go bad, no biggie. Every time PG has to remove something like this, it loses credibility. It means that someone’s mom or grandpa that finally adopted a tool at the urging of their privacy-minded relative will never trust that relative on this topic again. I’d rather see a few stable suggestions than a revolving door of collapsing products. To an extent this is unavoidable, but something like Skiff brought nothing to the table that wasn’t already being met by other options. And as other posters have pointed out, what it did bring to the table (eg. 100 gigs for free) was actually impossible.
5 Likes
I’ve worked at several P/E backed companies and know that “VC” doesn’t automatically = “bad.”
Rather than evaluate a company on its marketing, financials, technicals or products ask who are the people behind it? What’s their history, reputation, track record and general approach?
There’s very generally 2 types with some overlap.
- Angel investors using mostly they’re own money or money raised from other individuals.
- Vulture capitalists who depend almost entirely on issuing non-investment grade debt aka “junk”
Drilling down reveals 3 more types of P/E firms with the first 2 being beneficial to the ecosystem.
- Those who perform a healthy Darwinian culling of the herd.
- Those who efficiently squeeze every last bit of juice from the crop.
- Rapacious bottom-feeders who saddle their prey with backbreaking debt so they can extract all the equity and leave an empty shell to toss onto the bonfire of bankruptcy.
1 Like
3 Likes
One thing that struck me at the time was that it wasn’t just the dev that was hyper defensive, it was some of the posters here too. The longer the main thread went on, the more pushy those posters would get too.
I get the dev being pushy, he wanted to cash out. I don’t get people being impatient with the team for not adding them to the site. It’s not like they had anything to gain.
Many were likely skiff users themselves and wanted PG to validate their decision
4 Likes
I wonder how helpful PG’s recommendation actually was to the sale of Skiff. It seems like the buyer isn’t really interested in the current customers. But maybe other potential buyers were?
A company’s strong interest in being listed on PG isn’t necessarily a bad thing-- maybe mostly good. It really depends on what their future plans are-- which isn’t a criterion. As has been mentioned, having companies carefully align their services to PG’s criteria gives PG a lot of influence on the direction of that service and perhaps the industry. That’s remarkable.
Although in this case the enthusiastic alignment likely was a result of an intention to sell, I’d say that alignment to PG criteria is mostly positive.
It’s pretty annoying to have to switch email providers, so this thread has definitely been thought-provoking and worth thinking through. A few questions for the sake of conversation:
- To what extent does having a PG recommendation make it easier for a company to be acquired?
- How realistic would it be to include a criterion that explicitly asks the service provider to commit to a duration of their service?
- How hard would it be, on a case-by-case basis, to estimate how valuable a PG recommendation is to a service provider?
- How feasible would it be to ask service providers to enter into a formal agreement with PG which includes consequences for being acquired by a non-recommendable company (or other changes to the service)?
- How could PG adjust its communication to discourage a non-recommendable company from acquiring a recommended provider? For example, what if PG made it known in advance what it would do if that were to happen (e.g. send an alert to all PG forum users, post lots of information about why the acquiring company can’t be recommended, etc)
1 Like
1., 3. Privacy sells. Otherwise, there’s no reason to use these small tech companies instead of more reliable, well-established big-tech companies. As we see in this case, Skiff raised itself through privacy aspect of the service, then when it’s on the radar, it sold. It could be Notion or anyone.
2., 4., 5. Not possible. Just some weeks ago, Skiff didn’t feel like it would sell the business. No business would care regardless. They got the money. It’s time to go home.
The real solution IMO is to focus more on decentralized services.
5 Likes
Agreed.
3 Likes
If there’s agreement on this, how can the concept of decentralized be incorporated in the criteria? Could that be included in the Trust section?
The Best-Case criteria for Real-time Communication includes it (but it seems like it could be better reflected in the recommended services).
Should be decentralized, i.e. federated or P2P.
How could this concept be incorporated on the Email Service recommendations as well?
I have played with Mailchain and Dmail, which are the 2 main decentralized email providers, for a while now. It’s pretty much bare bone. And I am not sure about their business model. I don’t want to change my email often. In the end, I used Skiff, but obviously, I now moved to Proton and call it a day 
The recommendations page could more forcefully recommend using a custom domain name
1 Like
Well, there are privacy drawbacks to using a personal domain name to consider, which we note in the email aliasing section.
5 Likes
I think that the type of decentralization alluded to in the real-time communication recommendations is already the default with e-mail. E-mail is actually a commonly cited example of a (semi) successful decentralized/federated communications system.
2 Likes
That only goes up really for aliases. For your primary email that you give out to people you know it really doesn’t change things, right?
For aliases I am definitely agreeing as you know. But I think one should segment that between organisations that know your real name and are important ( like gov, bank, insurance, flights, etc) and less important stuff where your name isn’t necessarily known.
There seems to be a couple approaches emerging from the discussion: avoiding a recommended service becoming non-recommendable and mitigating the impacts on users when that happens. Both seem valuable.
Avoiding the loss of a recommended service
I don’t understand the effects of VC funding or the normal sale of private companies well enough to know which is more likely to happen and which is more likely to cause trouble for users of PG recommendations, but it seems to be worth trying as a Best-Case criterion (sorry if I’m using the wrong terminology) or at least addressing it for each service provider on the recommendation page.
The Email Services recommendations page reviews the Account Termination of each service. How feasible would it be to add information about what assurances the company can make to its users if it’s acquired? Ideally, it would be stated in their Terms formally and considered in the acquisition process, but if not, we could ask them.
Acquisition isn’t the only thing that might cause a recommended service to be lost or become non-recommendable. How could we address that too?
Mitigating the effects of potential service loss
Although what happened with Skiff might have a feeling of betrayal that we’d like to avoid, the concrete impact is having to change service providers-- especially unexpectedly. The approach of decentralization mentioned by @archerallstars and others seems to recognize that a recommended service will eventually become non-recommendable or cease to exist. In the case of email, the previous conversation has highlighted what might be the toughest impact: needing to change one’s email address domain with all of one’s contacts. As has been mentioned owning a domain and using an alias provider like addy.io or simplelogin helps, but there are also risks or trade-offs with those.
Could the recommendations page include a section about the impacts of having to migrate service providers? Whether it’s due to a sale, change or simply a new service with a different appeal, migration is essentially inevitable. Planning at the time of initiating a new service helps. I’m glad the Email Services recommendation page already includes the phrase “…maintain their agency from the service, should it turn bad or be acquired by another company which doesn’t prioritize privacy”. Could we have a section of the page that addresses this kind of scenario?
Service provider diversity
When we get an email from our service provider telling us about changes that will prevent us from continuing to use that service, it really helps when there is another established recommended service provider already in-place. (As we saw for Email Services, it can take a while for all the conversation to happen to become recommended.) I’m glad PG usually doesn’t just recommend one service provider. It helps for these scenarios when users might need to migrate. In some cases, it might also help with separating personas. In addition, it also seems common for differences to exist between service providers, whether it’s different features, better UI, support for a particular OS/platform, based in a country that particular users prefer etc. Despite the risks of adding new services, having multiple recommended service providers seems to have significant advantages.
2 Likes
Going back to what dngray said on the old thread(Remove Skiff - #80 by dngray)
This got me wondering about how this would have been handled if Skiff HADN’T announced that they were closing, just that they were being bought. I think they would have eventually been removed anyway, but I honestly think in that what-if scenario, I would have suggested that there be additional advisories being posted on PG saying people should migrate . I wouldn’t trust Notion in that scenario, even if emails are E2E encrypted there are still the possibilities of injected JS or stuff like that.