Communication and Secrecy at Serif

[drkanukie]

6 hours ago, debraspicher said:

In case anyone was curious on the subject of BC... I'm having a hard time remembering if anything was said regarding this or not, but just in case... sometimes their Twitter drops more info than is seen here in the replies... just depends.

 

The BC thing really grates, I have machines locked on V1 due to OS hardware restrictions and not being able to interoperate V1 <-> V2 is really painful. Given that they support PSD import and the affinity file format is interoperable between three different apps with different capabilities, I don't think this can be a technical decision rather than a marketing one to force upgrade revenue.

[loukash]

25 minutes ago, drkanukie said:

the affinity file format is interoperable between three different apps with different capabilities

Yes, because technically, it's only one app in three different incarnations, each of them having certain features disabled. That's why the "one document format" works.

27 minutes ago, drkanukie said:

I don't think this can be a technical decision

It's been confirmed by staff multiple times that the major changes between v1 and v2 were under the hood – i.e. the underlying code – so that they required a change to the document format as well. That's highly unfortunate and annoying – for me as well – but sometimes inevitable.
And that's nothing new in software development, and usually it has not much to do with "marketing", rather then with implementing new technologies.

[Corgi]

2 hours ago, loukash said:

...It's been confirmed by staff multiple times that the major changes between v1 and v2 were under the hood – i.e. the underlying code – so that they required a change to the document format as well. That's highly unfortunate and annoying – for me as well – but sometimes inevitable.
And that's nothing new in software development, and usually it has not much to do with "marketing", rather then with implementing new technologies.

I believe the point wasn't that v2 files should be able to be opened in v1, but rather that v2 files should be able to be exported as v1-compatible files, possibly with some loss of fidelity or function for any v2-only features since they aren't supported by v1 or its file format.

The proof is that v2 can export to PSD, and v1 can import to PSD, which demonstrates that there is at least a viable path. The notion that Affinity can export (with limitations) to PSD and other formats it doesn't own but cannot export to its own v1 format (with limitations) is absurd. I hope it's a feature that's in the works.

[Corgi]

Here's my take on how Serif performs with various aspects of communication. There may be other categories; this is off the top of my head. Feel free to render your rating on any of these:

1.    Telling us what's coming up short-term  [Fair, when there are active Betas, Poor otherwise]
2.    Telling us what's coming up long-term [Broad-brush: Poor    Detailed: Poor]
3.    Telling us what's in discussion within Serif short-term in response to crises [Poor or Late on some things, but possibly fine on others; small sample size]
4.    Explaining policies [Late, but Fair]
5.    Reaction to bugs on forums, with workarounds [Excellent]
6.    Reacting to questions of "how to use", with workarounds as necessary [Excellent]
7.    Communicating to gather information for use in product roadmap and feature prioritization  [Unknown -- suspect poor, based on v2 experience]

Regarding number 7: If I were a Serif employee taking forum feedback to heart, I would be pushing to perform a user survey. Every paying v2 account would have one entry in the survey, but there would be a second survey for non-customers (although since the latter is less controlled, its results would be less trustworthy).

The survey would seek to uncover how customers use the product, what features would be most valuable, what they like/dislike about being an Affinity user (product, support, policies, pricing...), and what suggestions they have for the product line and the company. The non-customer survey could seek to learn what's preventing the prospective buyer from buying the Affinity suite.

[Guest]

17 minutes ago, Corgi said:

Regarding number 7: If I were a Serif employee taking forum feedback to heart, I would be pushing to perform a user survey. Every paying v2 account would have one entry in the survey, but there would be a second survey for non-customers (although since the latter is less controlled, its results would be less trustworthy).

Surveys and focus groups are terrible ways of designing products, as the sample size is never truly reflective of the entire customer base (how many users will actually respond to a survey, are the questions phrased correctly, are they leading, are they meaningful), and group dynamics typically ensures that opinions of a vocal minority will influence the less vocal members towards a consensus that may not accurately reflect their opinions. They can both be useful tools, but rarely within product design/development. Serif probably already have a pretty good idea of what users want/have been asking for from these forums over the past 8 years, but actually prioritizing that feedback and putting into production seems to be the real blocker—which I suspect lies within management.

User testing of new concepts and designs with customers tends to work really well, even with small sample sizes—especially when done early in the production cycle. Not only can you test out your ideas and receive valuable feedback early, you will often gain insight into the mental models customers already have about various aspects of your products. That understanding is gold, and will often be found repeatedly as you continue to talk to your customers.

[Corgi]

1 minute ago, Bryan Rieger said:

Surveys and focus groups are terrible ways of designing products, as the sample size is never truly reflective of the entire customer base (how many users will actually respond to a survey, are the questions phrased correctly, are they leading, are they meaningful), and group dynamics typically ensures that opinions of a vocal minority will influence the less vocal members towards a consensus that may not accurately reflect their opinions. They can both be useful tools, but rarely within product design/development. Serif probably already have a pretty good idea of what users want/have been asking for from these forums over the past 8 years, but actually prioritizing that feedback and putting into production seems to be the real blocker—which I suspect is in management.

User testing of new concepts and designs with customers tends to work really well, even with small sample sizes—especially when done early in the production cycle.

Surveys are imperfect but using the forums as the primary (only) guide seems even worse (partially because of the reason you suggested -- it's hard to gather and prioritize forum feedback). And anyone who takes the time to register and post in a forum is also likely to respond to a well-designed survey (especially if an incentive for participation is offered).

I'm unclear on what you mean by "user testing of new concepts and designs" as a solution, since the problem is: how does the company know what new concepts/designs to implement and test? I agree that it's a good thing to do, but using it as a way to guide major feature development will be burning engineering effort implementing features on speculation that they're going to test well.

[debraspicher]

As much as I hate to say it, they do probably get plenty of feedback (and hyper-analyzing of things...). It's just the nature of designers. It's that they have to learn through actual practice pushing updates and seeing how customers actually use it. What is/isn't a hindrance, etc. That's probably the most valuable feedback... again I suspect developer reigns supreme when any plans are being put into the works as being focused on this in the pipeline and wanting to get those things nailed down flush, versus (designer) no this is not polished and we need to match certain standards, so we must address this now before this gets further down the assembly line... also, most who are actually using the programs giving feedback are generally the latter and do not comprehend the developer mindset at all, so they aren't helpful in that aspect... "it just needs to work", etc... so two types of feedback that are a total contradiction...

(edit) And tbf, they should address already paying customer's needs if something isn't functioning correctly first as well as polish. Versus adding new features that are much more likely to attract new customers...

[Guest]

3 minutes ago, Corgi said:

I'm unclear on what you mean by "user testing of new concepts and designs" as a solution, since the problem is: how does the company know what new concepts/designs to implement and test?

Again, Serif probably have a good idea of what customers are wanting, and their product managers/owners should be triaging these requests in their product roadmaps. Testing early and often allows teams to ensure that they are delivering the right features at the right time. It's not uncommon to test one feature, and have a user talk about how they would never use this feature, and would rather that you focus on something less feel is missing or poorly implemented. As this pattern repeats you get a pretty good idea of where you need to be placing your development efforts.

A user test can be as simple as sitting down with a few customers individually and showing them (still images) of proposed new features and listening to their feedback, or full on early (but largely useable) development builds. This way the design and development teams can gauge whether or not their assumptions are correct and have the opportunity to adjust the design of this feature before development has even begun.

[CM0]

39 minutes ago, Bryan Rieger said:

Surveys and focus groups are terrible ways of designing products, as the sample size is never truly reflective of the entire customer base (how many users will actually respond to a survey, are the questions phrased correctly, are they leading, are they meaningful), and group dynamics typically ensures that opinions of a vocal minority will influence the less vocal members towards a consensus that may not accurately reflect their opinions. They can both be useful tools, but rarely within product design/development. Serif probably already have a pretty good idea of what users want/have been asking for from these forums over the past 8 years, but actually prioritizing that feedback and putting into production seems to be the real blocker—which I suspect lies within management.

User testing of new concepts and designs with customers tends to work really well, even with small sample sizes—especially when done early in the production cycle. Not only can you test out your ideas and receive valuable feedback early, you will often gain insight into the mental models customers already have about various aspects of your products. That understanding is gold, and will often be found repeatedly as you continue to talk to your customers.

Yes, I agree on the user input, user input is an essential part of Agile development which ensures the customers have high bandwidth communication to the dev team as all features are developed alongside user feedback.  Iterative development, where the users get to experience a feature during the development cycle and influence the end result.

However, disagree on the forum.  It is simply not a good tool to be able to understand beyond the very problem you stated.  The most vocal will be that which rises to the top of awareness.  This is because the forum can't be organized, ranked or prioritized in any meaningful manner.  Things get lost in the forum as evidenced by mods who sometimes respond, "let me ping a dev again on that topic".  There is no systematic way to understand exactly how many customers are impacted by an issue or want a particular feature.

[Guest]

4 minutes ago, CM0 said:

However, disagree on the forum.  It is simply not a good tool to be able to understand beyond the very problem you stated.  The most vocal will be that which rises to the top of awareness.  This is because the forum can't be organized, ranked or prioritized in any meaningful manner.  Things get lost in the forum as evidenced by mods who sometimes respond, "let me ping a dev again on that topic".  There is no systematic way to understand exactly how many customers are impacted by an issue or want a particular feature.

I don't recall saying to use the forum as the only way to gather and prioritize what users are asking for. To me, this forum is one big focus group, representing only a tiny sample of the overall Affinity customer base. The big question should be how are Serif engaging with the rest of their customer base to better understand their needs?

[Corgi]

I guess I will have to just "agree to disagree" with those who believe surveys are unhelpful in gathering user input to aid in roadmap development.

Regarding forum feedback: I've done my share of forum reading these past few weeks, and honestly couldn't tell you whether the customer base is happy with the chosen v2 feature set or not. I'm sure Serif would have a better grip on this than I do (especially since they must get plenty of DMs). But still, I'm curious whether they view the v2 feature set as well-received or not.

[debraspicher]

The forums/social can be very beneficial for getting the feel of a user base. Obviously ignoring the "I'm loud see me roar" attention-seeking behavior and listen for the gist of most things coming in being said versus focusing in on specific "talking points" (for lack of a better term). I'm mainly talking about the negative form of these as they are very easy to become contagious among users, especially when it's added in alongside critique about how it's "not up to standard", even though it's not impacting their user all that much and they are still getting all the use possible out of the programs. The comments that should be concerning are when they can't get the usage that is advertised. This is powerful feedback.

Things that are very subjective (UI icon styles) versus usability issues. It's important to separate the two. I don't mind that Serif goes for a more classic look. I care that it's cohesive and my skills are transferable from other programs into the new setup regardless of the interface.

If I'm in the role of someone who wants my product to set a standard, if I read more than a few complaints about usability... it's in my list of things to look into first and foremost. It's high priority. It's not ignored. No reason to sort that in some sort of other schema.

[debraspicher]

10 minutes ago, Corgi said:

I guess I will have to just "agree to disagree" with those who believe surveys are unhelpful in gathering user input to aid in roadmap development.

Regarding forum feedback: I've done my share of forum reading these past few weeks, and honestly couldn't tell you whether the customer base is happy with the chosen v2 feature set or not. I'm sure Serif would have a better grip on this than I do (especially since they must get plenty of DMs). But still, I'm curious whether they view the v2 feature set as well-received or not.

I don't see how surveys can hurt. I missed the one that rolled through during V2 dev* because of childcare so it got buried in with the rest of my mail. I think if the developers are targeting something very specific in terms of an issue, then surveys can be quite helpful (more like a poll really)... if it's too general, then I can see where the results can be skewed by demographics and thus be more misleading...

[Old Bruce]

Regarding surveys: There is the, perhaps apocryphal, story that the Ford Edsel was designed to fulfill some rather extensive surveys. For those too young to remember the Edsel it was one of the worst sellers in automotive history. All I recall about it was the pushbutton automatic transmission.

[Corgi]

7 minutes ago, Old Bruce said:

Regarding surveys: There is the, perhaps apocryphal, story that the Ford Edsel was designed to fulfill some rather extensive surveys. For those too young to remember the Edsel it was one of the worst sellers in automotive history. All I recall about it was the pushbutton automatic transmission.

To be fair, the Edsel surveys were implemented using an extremely early, alpha version of SurveyMonkey, and the results were corrupted by a nasty VisiCalc bug.

[captain_slocum]

On 11/27/2022 at 6:33 PM, Patrick Connor said:

We reply to questions about how the software is meant to work and about bugs. We do not tend to respond to improvement requests. The software is not yet designed to do those things so it's not incorrect so much as not included yet. So no I cannot "set your mind at rest", sorry.

This thread has developed into an interesting philosophical discussion about the way a company is run, about tokenism ("Your custom is important to us") and about the way, as companies grow larger, they often become self-serving rather than customer led.

Patrick, I am sure your heart is in the right place, and I have no personal beef with you I assure you, but your reply is very telling. In my posts I have given you every opportunity to say something like "thank you for your considered feedback, your remarks will be passed to our development team for consideration" but presumably that is not how it works with Serif. This is how the whole thread started, with bemused Affinity users wondering, why the secrecy, why no roadmap, why no apparent acknowledgement of our suggestions. It just seems a pity to me that you (the company) appear to be wasting an enormous asset - a group of real-world users who care about your products.