Ask HN: What would make a great customer support tool?

20 points by prateekdayal ↗ HN
Hi,

I keep seeing posts on HN about customer support (email support) tools. We have been using Fogbugz for a couple of years now and I know that there can be a better solution. I have looked at a few other solutions (for example zendesk and RT) but I am wondering what a great customer support tool for a small team should be like. I have a few ideas

* Ability to reply/close/assign and do other stuff through email (RT does some of this but its extremely hard to setup)

* A very simple and usable web and mobile interface with a clean API

* Pricing based on number of tickets and not number of users (not all developers are support agents but you still want a login for them to see/respond to some tickets)

* Hooks to provide more information about the user, making it easier to provide a more personalized response. Basically making it easier to wow people who write to you.

I feel a lot of solutions out there are pretty complex (in terms of features, integrations etc) and quite costly for small teams. Do you think there is scope for a new product here? What would you want to see in it? I am going to start working on something very soon and would appreciate any feedback or ideas. Also, would love to find some beta testers :)

Thanks!

18 comments

[ 4.5 ms ] story [ 62.3 ms ] thread
I was discussing a new kind of support system with my friend the other day. We have been trying to understand current limitations of the support systems and if they can be fixed with technology (not always).

I'm looking forward to see some comments from people that are more into this topic!

I feel that a lot of the points made by the original poster are just bolt on features and not actually evolving support systems. Sure, they are nice improvements, but I think I've seen them all at one time or another. Like you said, not all problems can be fixed with technology. At least, not with the base email support systems we have now.

I'm more interested in seeing some out of the box thinking.

For example, build in some AI to preselect/recommend a base response from a pool of previous responses. That doesn't mean it'll be perfect (and still needs to be personalized), but it's better than starting from scratch or scraping through your wiki looking for a previous response.

Also, I want a view that watches an email support queue, public forums and possibly chats. That way my support reps can take the next available request no matter where it comes from.

Great feedback and suggestions. Would it be possible to talk to you more on this offline? If yes, how do I reach you?
I'm not thinking you meant me, but there are a few comments above who've not got profiles with contact details. You didn't say who you called out to! :-)

Just a thought! (as someone who finds nested threads difficult to read because of funky word - wrap and believes a shaded color bar is the solution to srting replies from OP / sub posters and has never seen it on any forum) 'nuf of my griping, but shout out who you mean!

Yesterday we had a discussion and following point we made: - nothing that we can do (simpler UI, faster search for the knowledge base etc.) is good enough to take it to market. People simply will not convert to our app because it will not fundamentally change the way they work - and help them fix their problems. We will still do the same thing, only a little bit better then competitors.

Also we joked that if a company want to make a user happy we can setup an engine that will send emails randomly from around 100 different templates assuring user that people are working on his case ;)

You say no-one will convery because your app is not fundamentally changing how they work.

How about you get really intimate with what they use, how they use it, and work on simplifying workflow UI, readability, all the human factors stuff which seems to be overlooked once a project gets rolled out because the painful bit of codifying processes too all the budget?

I mean, work out what you could overlay as a benefit, in a drop - in app. Ignore bling, save where used to flag data that needs working on urgently, and think about that old hullabaloo of the 70's consultant with clipboard: time and motion. Think how many times a UI context switch is required. Take away frustrations.

Obviously i've no idea if what yu're working on is the slightest bit suitable fr this treatment. But if i had a nickel for every time i've seen a support rep slam down their phone (i was going to say smashed, but we're all enlightened PC people now!) because of something as dumb as having to hit the fourth button on their machine to transfer a call, i'd be doing pretty good.

It's not fair of me to keep using phone use analogies, but i don't se enough info as to what other people ar trying to solve here, hope you get my drift anyhow.

Can you write "helper app"? Remember multifinder or i forget the name but whatever it was on early macs which allowed you to copy and paste to a stack? Id there something functionally simple like that you can offer which doesn't mess with the underlying interfaces, but which still requires some pretty neat writing and likely a fairly deep understanding of the existing apps to make work consistently and without snarling anything?

Whenever i get private project time, increasingly rare sadly, i keep sketching out UI models for a desk phone which would be complimentary to the tracking / contact apps i use, not be a awful ugly alterntive. Things like flipping a contact window from the list on the handset to my main monitor. The hard bit isn't so much now i have to work out a way to communicate that from phone to PC, because one way would be to loop the phone data to the PC and use the new USB screen driver tech. The hard bit is working out what you want to really happen. Is that, you finger flip a contact towards your main screen, and you get a fully expanded list of the discussions you've had with them? Can drag and drop work with this in a kind of wacom tablet way, so then hw does movement on the obviously smaller touchscreen translate to your main monitor? If, say, i've a preference for greyscale display on anything which cannot be usefully colored to convey real information, does the "flipped" window stay greyscale? Do i want the flipped window to be large, float or be even a normal window, or do i want it "brought back" / closed by a gesture on the phone screen? How could i do that from a cell phone?

That's pretty pie in the sky for a solo project, but whilst there remains good utility from desk bound objects (my favourite gripe is i can dial important customer numbers from memory faster than i can bring their entry up on my very latest processor Nokia) and these objects, like everything else, wear out, and are now nicely converging to voip protocols which are mature enough there's no reason to go for a proprietary type such as cisco, theres a market for these things.

Making any of this elegant is no small ambition. But consider this thought, that i remember new employees getting "phone training" because they needed to know how to handle the butt ugly interfaces of pbx attached phones which to this day are not often much better than a bunch of soft keys and if you're lucky those keys are customisable but still very basic features. Why can't i drag and drop a call with my finger whilst i'm on another call? The other thing is with the iphone we really have a lot of people finding new ways to use a telephone, and hat idiom i think will spread to the desk phone. Suddenly it may not be necessary to train someone how to put a call on hold, transfer and pick up the next incoming without muffing it with fat f...

We have just set up tenderapp for our new Beta site www.pinkelstar.com (easy social network integration for iPhone and Android). I like it a lot so far. However, since we've just gone live I can't yet say anything about daily use. But the setup as pretty easy and convenient and the price was pretty decent too.
Seconding Tender. Beautiful and simple.
Maybe just my mass of plugins but FF3.6 isn't loading tenderapp, while chrome just fine. s'pose could be google's search redirect playing funny with requestpolicy, but am confused :~/

Apart from pricing, what's better than FogBugz, in terms of your decision? I can't find a straight features page, though straight to API faq so that's good.Seems to be Ruby, whilst Fog has XML and presumably, being .net a few other interfaces.

What confuses me is that OP wanted or hinted at a desire for a expanded feature set that Fog doesn't apparently have. Which are they that Tender gives? (I got the API point, that's good, but Fog being based on the CLR would give more ways to extend it if you took the site licensed version instead of the hosted version.

Also nitpicking, but Fog for the site install prices much more nicely than the on-demand version if you've >5 users.

I'm just utterly confused as to what was the OP's question was - it's so utterly vague - and led me to ramble off in different directions above. So what's the comparative benefit?

I'm a fan of Joel Spolsky's writing, and have evaluated Fog online, from POV of whether it would be useable by non programmer staff as comfortably as i'd like.

but i'm not yet a customer as you might guess from my above psts i've interest to do things which aren't standard for a code shop. Basically, i know the value i get is in much deeper planning and i'd have to write a lot of what's desired by my company, meanwhile Fog looks robust enough to use it as a way of pulling together a few things which aren;t totally out of its design intent. Outlook integration is a useful thing not so much for management but for syncing nicely with the cell phones we use, e.g.

I mean i'm not shilling for Fog, i just don't know what the criticism is except price. What is the OP disliking about the Fog APIs e.g.? If i put my cynical hat on, i'd just say this has been a "Ask /." kind of experience. I'm interested enough in the subject to engage, but have no way of really replying. Was the Q really "what's not Fogbugz out there are cheaper?" I genuinely don't know, nor can i (presently) find any links to older HN discussion which might be useful.

All this management software, whether for bug tracking or whether you make it front of house or anywhere in between is so useful to any helf - evolved business of any size, i seem to remember IBM paying a pretty penny to get ahold of Lotus who sussed the business pitch in the 90s.

I suppose that what i'd like is a discussion as to what's state of the art, what companies or systems are really pushing the limits and getting things good and affordable, not a mere comparison with FogCreek which is very well written up for a company of any size.

I'll post separately with my confusion points, so please understand i'm probably just missing the point here.

best to all.

Thanks for pointing out Tenderapp. I'm using FogBugz, but this looks like it might be much better for us.
I've thought a lot about this and sketched out some ideas myself. I work at apartmenttherapy.com -- which is primarily a content/advertising oriented business, though we are building toward some more app & community-oriented features in the near future. Most of the systems I've evaluated tend to fall into either a "bug tracker," "help desk" and/or "CRM" bucket, none of which is a good fit for what we need. What we need is more of a "reader communication" platform. We get a LOT of incoming email but we don't have very good tools for really tracking it. Options I've evaluated are generally too rigid and expensive, or flexible but a lot of work to adapt & maintain , and ugly (e.g., RT).

Being able to manage the system solely over email, at least for basic interactions, I think is really important for my team. I really liked some of the features of Email Center Pro (http://www.emailcenterpro.com/features/) but found their UI a bit too clunky and the inability to do basic management directly in the email client was a killer. A streamlined/updated version of that plus email integration similar to RT would work really well for us, I think. If you want to chat more, please email me scott AT apartmenttherapy.com.

And I do have a bunch of potentially interesting feature ideas: things like "mood" analysis (based on the text, does this person seem upset/angry/happy/etc); hooks into custom/proprietary systems (is this person a paying customer for product X); analysis (is volume of email type X higher than usual? Might want to look into it, you may have broke something / made something awesomer).
One of the ideas was to create something like a local business social network and use it for connecting employees and customers. You would ask your users to login with their Facebook account to your system (benefit - a lot of info about the customer) and then hook them up to the product they have and deal value they generated. This would give you good info about importance of the call.

Idea could be summed up as "Quora for support with CRM integration".

I think the problem with Facebook, is in a straight B2b job, quite apart from the fact i think i could never legally clear working with FB apis, is the info people put on their FB is not relevant to their work. The info i want to plug in to is such as reporting structure, tenure in a position, line responsibility, projects, published papers, conferences attended or talks given. That's very specific stuff which sometimes is publicly available, but usually not. THis intel comes off the direct sales floor. But it can be super useful when you need to work with a customer to understand the scope of their needs, because you have some kind of map based on which you can (if you are good at this kind of interpersonal) start asking how your bit of code affects group a in a related department, or if a change to say a account presentation will go down well with CFO. There are commercial databases which try to keep up with personnel changes in bigger companies. Thomson does one. "Amadeus" a Belgian data source does well for company reporting structure. I'm a firm believer in selling features up the chain in a customer site, not "upselling" them new products. Dones well, this has sometimes had for me the side - benefit of locking consultants out of the loop. Nothing worse than aiming a project or sub - project at needs you've discovered at a custmer, to have a different management branch bring in a consultant to "evaluate" your work's worth. So, i try to insist that the primary role of sales is not necessarily to be closing deals, but to be closing the data we need to work quickly with a customer. I'm specific too, we never use the word "client" beause it sounds too open ended, too "maxing out billable hours". I try to teach sales that the efficacy of their sale is directly related to how easily we can start to work as closely as is practical or legally appropriate with a customer. Basically, play the consultants at their own game. You can bet they know who is who in a attractive potential customer site, because, generalizing,that sems to be the only leverage they ever have.

It's a nice idea you have there. But it's 20 years since i first experienced customer receptions declining to give any info as to employees, even for specific roles / names. I've no idea how legal in a big corp would think, about even using FB personal data to approach or manage contacts with employees. That said, the first thing i learned about sales, which is still an essential component of your job, i'd say particularly so if you're coding to fix a customer problem or desire, is to verbally map out common contacts with the first person you speak with. That's not far from what FB actually does, but you're doing it in person, albeit down a phone call often as not, and so concerns as to privacy can be finely guaged if you learn to listen.

The kind of data we pull in for any company we'd like to wok with is theses, lectures, of who we think pays attention to what we do, even if not "decision maker" then we delve deeper to try to understand the scope of a department influence - are they a skunkworks, with board support, are they mavericks, are they hard pressed ops guys? Sometimes just because who you communicate with might be classified as one of those types, you may be doing a lot of help by widening your contacts outside, because for big corps it's not a given teams talk to teams talk to the right management layers all the time.

What i'm on about i think works for any company you work with above 50 employees, and requires a lot of discretion. I don't have any "hard sell" function in where i work, but it's good to learn off experienced sales guys, and quite often they've worked "hard sell" before. My definition of "hard sell" is to just be absolutely direct about what you want to convey, waste the least time of any party, and haggle price especially if your bit of the project can give longer term or wider benefits. My own experience is you're lucky if you an pitch a technical project at a senior manager and get two way conversation, so the idea is you mix...

There's definitely room for such a product... which is why I formed a company with a friend and wrote one. :)

Our product is currently in closed beta as we squash bugs, but we hit every bullet point on your feature list, plus a few more that we've uncovered as our customers use the system.

Of course, we use this for all of our bug-tracking and customer support as well, and it's amazing what having a good support system does for your workflow.

So, we're competitors, but don't let that stop you. After all, competition is the crucible in which truly amazing products are formed. :)

Thanks for the encouragement. Yep .. healthy competition is always great and exciting to have :)
Below is opinionated, but hopefully worth airing:

My view is that the field is over-populated.

Probelm is often definition of what's wanted. You end up with overlap between issue tracking and CRM, and if you add in a few outlier requirements you've described half the commercial software i see, from email installs to business automation, ERP and even EDI.

That's quite an exaggerated generalization of mine there, but the list of comercial and non-commercial software which can be put to use as you describe is enormous.

What are you really missing from FogBugz? I'm not clear why you're having to look elsewhere.

Mobile apps integration is simply not mature, or rather elegant, for any product i've seen. Zimbra, Lotus Sametime, come to mind as rather half hearted efforts requiring expensive additional licenses. Developer business sales groups really should wise up to the idea that a good mobile app is the thing small companies (mode 100 employees) i know want to have better systems, it's not a mere extra for new installs, it's the driver.

I forget when i first heard about "unified communications" but it wasn't even in the last decade. That promise seems to have been eternal. I feel we are inching closer, though, yet there's nothing breakout good in any category which i've encountered.

What i want is products with common rules engines, or something which can overlay that via apis. Smetimes i'd just like a regex to flag a string in an email as signifying a task is done and update a conversation thread. Sometimes i'd like a proper transaction written back to a sals DB. Sometimes i'd like to know my updates are delivered, with a message broker or transaction manager.It's applying consistency to that which makes me want a common interface to software as varied as voive, shceduling, management accounting (e.g. for that last one where there's a showstopper ticket, i want the revenue forecast to be dropped right away for that customer) Right now even trying for that flexibility requires some big bloat installs and a lot of work, some of it pretty low level.

As to scope for a new product, i don't think you can design a product out of the box for a large proportion of users in your situation. I believe ther attempt with XML and XSLT to find a common interface for data munging was a good try, but it turned out the only thing more verbose than the XML was the acronym dictionary. And a lot of XML i see is serializing ASN or whatever talks to the outside world of suppliers, basically a patch, not a consistent code base, so you are doing mapping schema for more than just your primary DB.

Sure, i'm being very opinionated / hand - waving. But maybe the best anser is to say you should think hard how important your support process is. Is it something which can really make you stand - out? If so, maybe you should invest in time and coding effort to get what you want. There really is no shortage of systems, apps and tools. Just nothing i see which can hook diverse office systems together. Example of this is getting desk phones to be useful. The best i've seen so far is the XMPP calls which SipExec and Aastra kit supports, but still you need some glue, and a lot of thought what you want delivered and why.

I'm emphasising voice in my post because no atter how i type, there are so many times a quick call really sorts out a problem in a fraction of the time. In my integration nirvana, i'd have the phone logs transcribed and appended automatically to the ticket. I'd have my desk phone call up the subject last discussed when i dial a number, preferably on the desk phone itself, so my concurrent work windows can be left alone.

If there's an opportunity to introduce a new product, i think the hurdle is the mind boggling variety of things any successful product would be wished to work with, and you'll bloat at the end of the day.

I doubt that what you want couldn't be done fairly simply on the data front. But UI on mobile devices is a good challenge to put it mildly.

Two favourite related moans of mine is the state of Nokia supp...

Ive been a fan of www.kayako.com for some time, its got lots of features and is relatively inexpensive.

For a new product what I would like to see: Easy Template System Simple Back-end 1 Click Install (simple to wordpress) Live Messaging

1 area that I do not feel has been explored or taken advantage of is desktop software. I tried a while ago to create a desktop app that would sit in the system tray, receiving status updates and messages from a central system.

It would allow customers to stay up to date with system status without visiting a website and it could be used to notify of support replies, which should then reduce the "I didnt get a reply" issue.