vBulletin --> Discourse


(Fahad Ullah) #1

Hello everyone,

My name is Fahad Ullah and I am from Lahore, Pakistan. I work at PakWheels.com as a Product Manager and one aspect of my job is to handle and grow the community of automotive enthusiasts from Pakistan.

At this point, I am contemplating on migrating our community from vBulletin to Discourse and I came across FeverBee while reading about similar migrations. I found the FeverBee Community really helpful and decided to make an account to learn from the experts and contribute wherever I can.

My Dilemma

A vast majority of internet users in Pakistan uses the internet for accessing Social Networking sites and are used to the kind user experience that these websites provide. In the last few years, we have noticed a significant decline in the activity on our forums. To investigate the problem, I spoke to the current users and some older members and got to know that they feel the forums are complicated, slow and very hard to navigate. They also feel that it is very hard to search and find information on the forums. Keeping that in mind, we decided to give our forums a new look and feel and move them on to Discourse from vBulletin.

One of the primary motivations to move the forums on to Discourse (apart from its feature set) is that we are a Rails-based Software House and our main website is built using RoR. All of our developers are well versed in RoR. vBulletin is practically dead and there is no point spending development hours on a platform that has a huge maintenance cost for us.

I have got the full approval from Head of Engineering and the top management on this move but I am a bit reluctant to pull the trigger. It will be a huge move for us as our site has close to 350k registered users, 4.7 million posts and 242k topics.

I am reluctant because, there is a huge cost involved and this is will be my first experience running a migration of this magnitude. What if even after the migration, things don’t go as expected and we will not be able to get the same engagement levels that we previously used to enjoy?

And if I do decide to pull the trigger, what things do I need to keep in mind from community point of view to make the transition as smooth as possible?

Looking forward to hear from the experts

  • Fahad

So, what are you working on?
(Sarah Hawk) #2

Welcome @fahadullah – I’m going to split this out into a topic of its own so it doesn’t get lost in the noise.

I hear you. I’ve been in exactly this position, with a migration of around the same size. A lot of this will probably sound familiar to you.

The first thing that is important to note (and I’m sure you already know this) is that a struggling community won’t be saved by a change in platform. If your current platform is lacking in functionality (a lack of functioning notifications, for example) then it will no doubt help, but it’s not a silver bullet. Be careful that you don’t make promised to management that you can’t keep.

That said, I support your decision. The experience that Discourse provides is vastly superior to that of vB.

A few things of note:

  • Have you analysed your traffic sources to ensure that the bulk of your users are on browsers that are supported by Discourse?

  • Be very careful with your redirects if you rely heavily on organic search traffic.

  • This is your opportunity to take a hard look at your content. Do you need to migrate all those user accounts and posts? I’d think about only migrating stuff that has been viewed in the last 3 years, and only active accounts.

  • This very much depends on the community, but I believe that communication is a very important part of this process. Your members likely feel some ownership over the community, and even though it’s your decision to make, keeping them in the loop will ensure that they do what they can to help you migrate smoothly. Don’t ask for feedback that you can’t act on, and remind people that the community isn’t a democracy, but do give them the opportunity to have their voices heard, and do what you can to alleviate their concerns.

If you have other questions or concerns, I’m happy to answer them.


(Nick Emmett) #3

Hi @fahadullah, whereas I don’t have any experience in either of the platforms you’re dealing with, or have I ever migrated platforms, but I can see from what you’re saying, it looks as though you’ve approached the early stages in the right way. and involved your community members. If you’ve got feedback from them and your current platform can no longer support you and them, and in fact becomes a drain on you and your resources, then it sounds like a shift in platform is a good move.

I totally agree with @HAWK in terms of making sure you’re not looking to move just for the sake of it and that a platform change will not save a struggling community - that would need a deeper look at your strategy rather than the platform. Setting expectations to both your members and your leadership team is vital to make sure you can deliver what you set out to.

good luck, keep us in the loop about how it works out, and make sure you come back if you need to talk things through more!


(christopher w) #4

Hi Fahad,

Given the size of your community, I would at the very least carry out the following due diligence before committing to any new platform:

  1. Take references from existing users of the platform you intend using. Obviously the references should be taken from folks running communities (if they exist) of a similar size to your own. If they don’t exist then point 2 takes on a particular significance.

  2. Load test the platform using an automated load test system. Test it at the levels you’d expect to be in 24 months given you cannot assume that the platform will improve its performance over time.

  3. Get a subset of your existing community (preferably influencers) to road test the new platform. Pay special attention to the good parts of the current experience and ensure the new platform caters for these just as well, if not better.

  4. Check out the platform road map and make sure it matches your own - at least in the most significant areas. Note re point 3 above, Discourse doesn’t, and doesn’t intend to, provide a WYSIWYG editing experience. Does that matter to your 350k user base?

  5. Involve the platform vendor in your plans. Note any concerns. Embrace efforts to assist you.

Apart from being able to demonstrate to the chain of command and user base that you’ve done your due diligence, the above will let you sleep a little easier at night and will highlight, and allow you the opportunity to fix, issues before your user base encounters them.

If it would help, check out my LinkedIn profile (details in my profile here) - you can see the types of Enterprise systems I’ve designed and worked on over the past 30 years - it might give you some comfort as to the validity of the above advice.

Good luck,

Chris


(Fahad Ullah) #5

Thank you @HAWK for the warm welcome and making this post a separate thread. [quote=“HAWK, post:2, topic:2180”]
I hear you. I’ve been in exactly this position, with a migration of around the same size. A lot of this will probably sound familiar to you.
[/quote]
I have read through all of Site Point case and even read the threads posted before, between and after the migration to read your teams’ and users’ experiences throughout the migration process.

I must say, those readings have really boosted my confidence in Discourse and this migration.[quote=“HAWK, post:2, topic:2180”]
The first thing that is important to note (and I’m sure you already know this) is that a struggling community won’t be saved by a change in platform. If your current platform is lacking in functionality (a lack of functioning notifications, for example) then it will no doubt help, but it’s not a silver bullet. Be careful that you don’t make promised to management that you can’t keep.
[/quote]

I am very well aware of this and I am first addressing the underlying issues with the community that I am facing (bad moderation, lack of dedicated community manager, forum organization, etc.) and then I will move onto the product related changes (which is my actual domain).

Our existing platform lack a lot of essential functionalities and building them on vBulletin will make things more complicated.

That is not an issue as more than 85% of our traffic is on the supported devices and browsers.

That is a major concern for us and our SEO team will be working with us on support throughout the migration duration.

What are the pitfall of migrating everything? Does this affect the performance in any way?

That is very important. I have taken the influencers of the community already in confidence and will run a private beta with the small fraction of the community before initiating the migration process.

Once again, thank you for a detailed reply and a warm welcome. :slightly_smiling:


(Fahad Ullah) #6

Thank you @Nick_Emmett. I will ensure the points mentioned and will address low hanging issues before doing something radical.


(Fahad Ullah) #7

I haven’t tested this but this is a very good suggestion. I will make sure that this is done before making any final decision. [quote=“ccdw, post:4, topic:2180”]
3- Get a subset of your existing community (preferably influencers) to road test the new platform. Pay special attention to the good parts of the current experience and ensure the new platform caters for these just as well, if not better.
[/quote]
This is already a planned activity and an instance with a fraction of data will be given to a set number of users to test and provide their feedback.

Being a Rails house and on an Open Source Platform give us a lot of liberty in building small features that are required. But you are spot on, on involving the platform vendor and communicating with them.

Thank you so much for the detailed reply. I will keep the things mentioned in mind :slightly_smiling:


(Sarah Hawk) #8

It doesn’t affect performance per se, although it took 4 full days to run our migration (280k users, ~5 million posts) and there was a lot of data clean up to do. My reason for asking is that it’s a good time to do a cull, esp of user accounts that aren’t active. There’s no point in compounding your technical debt with legacy data that isn’t relevant. If you think it IS all relevant then go for it.

And you’re welcome. I’m glad you’re feeling more confident. Keep us updated on your journey, and let us know if we can help along the way.


(Rob Bosch) #9

Welcome @fahadullah
I read this conv with a lot of interest. I am member of 2 communities which both use phpBB. I hope both will migrate to discourse too since the useability of discourse is so much better.
Fortunately those 2 communities are not that large as yours, but still it will be a challenge (first to convince and secondly to do the actual migration)


(Sarah Hawk) #10

As an aside to this, @ralphm has just joined us here. He was a senior member on my team of volunteer staff at SitePoint when we carried out the migration discussed above. He will likely have an interesting perspective on how the transition was for moderation staff.


(Ralph Mason) #11

Yes, definitely. The transition can be a bit of a shock. Moving from vBulletin, you first scramble around wondering how to replicate all your old moderation behavior. Then you realize there’s a whole new way to approach moderation, and you relax a bit and actually start to listen to what Discourse is offering. Overall, I feel things are a lot simpler now, even though Discourse moderation takes some getting used to. That was our experience, anyway. :slight_smile:


(Ophelie Lechat) #12

Another aspect to consider (from another SitePoint team member! There are many of us :slight_smile: ) is your traffic and monetization (ads) of the site, if those are important factors to your team.

Our traffic dropped quite a bit after the initial move. Engagement stayed about the same, and grew over time, but for an ad-monetized site (as yours seems to be) you need to carefully consider things like SEO, redirects, engagement, regular email, and so on. Some of these we considered, some we didn’t. We migrated a large chunk of our content over, and left the rest on the old site as an archive. Both traffic numbers combined were still much lower than our original visits.

The site refresh, the improved functionality and being able to use a RoR app made it worthwhile for us, but if you’ll need to show increased revenue and traffic numbers to qualify it a success, make sure you’re prepared for a few months of instability.


(Fahad Ullah) #13

Yeah. In that way it makes sense, but we rely on organic traffic a lot and sometimes we get engagement and replies on even 5 years old topics so I think I will have to carefully analyze what should be migrated and what should be left on legacy system.

Thank you @Rob_Bosch

I hope your communities’ admins move their sites to Discourse too :slight_smile:

Awesome. Let’s hear his thoughts on the moderation aspect. :slight_smile:

Welcome and thanks for your reply @ralphm. Could you please let me know exactly what kind of shock this would be from moderation perspective?

Definitely. Ads are important to us, but our forums exist as a platform for auto enthusiasts and generating revenue through them, although important, but is not a priority. However, organic traffic and engagement levels are very important in our case.

I am expecting things to take a hit and then gradually stabilize. If that remains the case then we are good I guess.

What was the reaction of the users? Was there any resistance from them? And how did they like Discourse initially after the migration and after some time? Did there opinions change over time?

Sorry for too many questions :slight_smile:


(Gear Buzz) #14

In my experience the usual choice for a migration would be for a VBulletin forum to move to Xenforo.

I have my eye on a move from my current VBulletin 3.8 to Xenforo but am waiting for Xenforo V2.0

But there is no fixed release date in sight for v2.0.

As I wait, more and more add ons are being created (like a classifieds add on) it is being re configured to make it (even) easier to develop for. I see this as good and there is no harm in waiting for it to move into maturity with more very useful and reliable add ons. It is still the new kid on the block. Also the longer I wait, the more developers there will be that have learned how do work with it.

I know I am a development Junkie. I want to be able to think up features and have my devs make new functionality live on the forum quickly. So ease to develop for is super important to me.

When Xenforo 2.0 is out I will reassess the platform landscape and if Discourse leads the pack will definitely consider it it I but presently I feel I am headed towards Xenforo.

The primary goal of V2 has been described as - making Xenforo easier to develop for.


(Ralph Mason) #15

“Shock” may have been a bit overdramatic, but the main point is that you need to rethink your approach to moderation with Discourse, as the tools are quite different. My memory of vB moderation is getting hazy now, but we had ways of doing things in that system, and the temptation was to try to mold our Discourse moderation to fit our old workflow, which wasn’t a good idea. You’ve got to explore the tools available in Discourse and try to understand their approach to moderation. The tools and their various uses are a bit arcane, unfortunately, so a lot has to be learned by trial and error, often with multiple personal accounts that you can experiment on.


(Gear Buzz) #16

Can you remember any of the differences?


(Ralph Mason) #17

The moderation tools are quite different—from how members report problem posts to how moderators deal with them. We had developed a somewhat elaborate workflow around the tools available in vB, and essentially tried to keep that workflow when we moved to Discourse, even though the moderation tools were quite different.

As an example, we had a moderators area in vB where we could discuss problem posts, and we tried to set up something similar in Discourse. It was more trouble than it was worth in Discourse.

Discourse also has a number of options for dealing with problem members, some of which didn’t match the options we were accustomed to. So at first we pretty much ignored the Discourse options and manually addressed issues ourselves, which was a real mess.


(Joe Velez) #18

I’m a vbulletin guy myself (aka Princeton).

We humans hate CHANGE.

We are talking about a MAJOR change here. vBulletin and Discourse are nothing alike.

You will have a huge BACKLASH in your hands.

The reason I say this is because…

It’s evident that you don’t make big changes to the vbulletin UI. So, any big change to how people use the site will not be welcomed.

Your audience may not be ready for such a major change. What have you done to prepare them for such a big change?


Your site and Sitepoint are not equals…

Sitepoint is known to constantly make changes to it’s forums (more than the average) - especially the forum “homepage”. They definitely made a lot of changes to the site as a whole in the past few years. They are constantly evolving and members now expect change as the norm.

The Sitepoint audience are more acceptable to change than the average group.


Change is inevitable. We need to evolve to grow. I’m glad you know this. :slight_smile:

It’s great that you ask your members for feedback but it’s important to know that sometimes what they say is not what they do. Review your analytics - take a close look at what your audience is doing.

What’s your best channel? Organic traffic? Social? Direct? Referral?

Focus on your best traffic channel. I’m not saying to ignore the rest. If Organic Traffic is your best channel then make sure that you don’t do anything to negatively affect this.

To increase activity/traffic, read the messages that members leave on your site (not about site feedback but in general) you may be surprised at what you find. Ideas are often found where you least expect them.

What are they looking for? What do they want/need? Satisfy their need(s).


To prepare your audience for a Discourse upgrade I recommend you start making incremental changes now. Get it as close to what you believe your Discourse installation will look like. Do this in a 6-12 month period. Before you know it, your members won’t even notice that you upgraded to a new system. Your members will be talking about the awesome new features – Not about the “ugly changes”.


If site is slow … You may want to look at a master/slave setup if you don’t already have one. (Personally, I found the speed ok.)


(Sarah Hawk) #19

No such thing. :slight_smile:


(Sarah Hawk) #20

I’m not sure that either of those points are wholly fair, but I’m very interested in your perspective, and I support your position about humans being naturally change averse.

I don’t think our audience were more accepting of change than most, at the time. The relaunch was the first of the changes, and I worked very hard to manage the process, which DID come as a shock to everyone, but it certainly wasn’t insurmountable, and was well worth it, almost immediately.

The thing that I learned fast was that you have to be firm, and while it’s important to listen to your audience, the community isn’t a democracy and that needs to be made clear.

And remember, that for every one person that vocally disparages the choices and changes that you’re making, there are 10 silent ones that support you.