I embarked on a year-long mission to find better project management software for Springload. Streamtime, which Springload has used for the last few years, just didn’t cut it. It wasn’t easy finding software that would help navigate the life cycle of a creative project while being relatively painless for everyone to use. Although I was almost paralysed by the fact that I didn’t want to make things worse (I felt responsible for going with Streamtime), I whittled an intimidating longlist of 43 maybes down to a shortlist of four. Here are a few notes from the journey.
What was wrong with Streamtime?
Streamtime. Sigh. Let’s start from the beginning. Springload originally used a chain of management tools: Excel for estimates; Word for proposals; a simple tool we developed ourselves for time tracking; and Quickbooks for accounting, reporting and invoicing. This actually worked pretty well, but as Springload grew, managing all these different tools became unwieldy; we went with Streamtime to centralise everything.
But Streamtime just had too many niggles to ignore, no matter how much we tried to work around them. There were some frustrating interface quirks, which undermined people’s confidence in it – especially since we’re a design firm heavily into user interaction. The infamous “window limbo” was the best example of this: pop-up windows refused to close unless you knew the one magic combination of clicks to make them go away.
Certain aspects of Streamtime were clunky or limiting. I’d hoped that the pared-down versions of Streamtime might insulate the majority of staff from using it any more than they needed to – but the mobile app didn’t do timesheets and Streamtimer, the Adobe Air desktop widget, was unusable. Using Streamtime as a desktop app (as opposed to Software as a Service) wasn’t ideal either – Streamtime uses Filemaker, and ensuring it was installed and kept up to date across multiple machines was a real pain.
On top of all that, Streamtime couldn’t cope with the number of projects and tasks that we have on the go on a continuous basis. Reporting was slow and inflexible. Job numbers were problematic. Numbers didn’t add up. The staff hated it.
In a presentation to staff about potential replacements for Streamtime, I opened with a slide that read “Farewell Streamtime – it’s not you, it’s us”. The next slide read, “Farewell Streamtime – actually it is you”. A joke, but one with a firm basis in truth.
Flexible features for an imperfect world
I knew that whatever we went with next, the most critical feature to get right was the timesheeting. No one likes doing that part of it, everyone has to do it, and for the majority of people, that’s actually the only thing they’re going to have to interact with. I wasn’t going to compromise on that – if it did great reporting, but the timesheeting experience was crap, then I quickly disregarded it.
It also had to be Software as a Service (SaaS), which makes much more sense – something that runs in any browser, anywhere, any time. There’s just so much less technical overhead when it’s hosted in one place and it’s accessible from all sorts of devices.
The core features – timesheeting, projects and tasks, invoicing, reporting – needed to interact in a meaningful and flexible way, particularly when you need to go back and make a change. Things go imperfectly often enough that there needs to be a way to keep track, but it was frustratingly difficult to find software that would acknowledge this. If, for example, some incorrectly added hours have to be removed from an invoice, when you adjust those details in the invoice I want those hours to go back to the status of unbilled billable hours. Lots of programs let me make changes to the invoice, but few then ensured that those changes were automatically reflected elsewhere.
In terms of oversight and reporting confidence, it was also important to find something that might highlight, for example, that we hadn’t billed someone, or that there were some hours sitting there that no one had invoiced because they got timesheeted into the wrong month. I wanted something that could really put managers at the helm of the ship and steer the whole operation, but too often the view was patchy at best.
A process of elimination: “fast fails” and marketing guff
There were a few sure “fast fails” that let me eliminate some prospects straight away, such as having no invoicing option, or being too ridiculously vast in scale. Often this was a relief, since it spared me the rigmarole of really putting the software through its paces. Doing the 30-day trial – creating a client, adding a project and adding tasks to that, adding dates, logging in as five different imaginary employees and timesheeting against those tasks – started to get me close, but there were deeper features that required consideration, such as how reporting worked across multiple projects, or how functional the invoicing was. So it took a year.
After a while I got pretty cynical about the online marketing guff written for prospective software – it was just a low frequency buzz in the background. After you’ve read some amazing marketing material and logged in to run through your trial, and then you spend a whole lot of time setting stuff up only to find it doesn’t do what it says on the can, you just give up on reading what’s on the can. I started to look for what they didn’t say, which ultimately was more telling. You start to think, hmmm, they haven’t mentioned invoicing – ah, it actually doesn’t do invoicing.
Confirmation bias – it needed to look at least bit pretty
Given that Springload is a design agency, the look and feel of the software wasn’t a part of the user experience that I could just discount – it was important. I felt that confirmation bias (the way people generally lean towards information that confirms their beliefs) would influence how much confidence people would have in it at the end of the day. If it looks really good, you believe it’s going to work, and you look for confirmation that’s the case; if it looks half-baked you assume that it is, and you look for stuff that confirms that. At least one candidate was swiftly shown the door because it was too ugly to fully assess. And one of the shortlist candidates almost got disqualified because of its shocking logo (WorkflowMax have updated their logo since then).
The ideal solution doesn’t exist (or is really, really hard to find)
Although the market for project management software is pretty crowded, I was surprised at the absence of any clear market leader or industry standard. The closest thing was Harvest, so that made it to the shortlist nearly by default – no one ever got fired for choosing Harvest.
Some features I coveted were difficult or impossible to find. I wanted decent CRM (Client Relationship Management) so that we could keep track of prospective clients and manage their contact details, but it wasn’t always offered. It was also tricky to find software with a hierarchy for tasks, such as that found in Microsoft Project – the ability to create relationships between tasks so when a change is made to a parent task (such as the start date) that change filters down to subtasks; or to create dependency amongst tasks, so you could require that one task has to be complete before another one can begin.
I gave up altogether on finding a feature that would offer an overview of available human resources, so that allocating time and spotting gaps was easier. Granted that’s a pretty big ask, looking across an organisation like ours – you’ve got multiple people in the same role, multiple designers, multiple front-end developers – but it would be nice to have that kind of oversight. Wishful thinking, maybe.
I’m open to the fact that transitioning to new software might mean we have to change the way we work a bit. We used to give all of our estimates as a range of time, but since that’s not really supported anywhere maybe we’ll timesheet the high-, low- or midpoint. I don’t want us to shy away from change if there’s opportunity for improvement.
The short list (and one honourable mention)
And then there were four. We got four teams at Springload to trial the shortlist candidates. It was going to be a tough call and there would be inevitable trade-offs – there was no clear, obvious winner – but I was hoping that my longwinded selection process would result in a solution that most people would be happy with. If you’re interested in seeing the long list, and the various notes I made along the way (some kinder than others), feel free to get in touch.
Tagline: “Job, time & invoice management software”. It’s just a bit clunky and hard to use – optimistic that Xero might fix that.
- comprehensive – covers everything we want
- bought by Xero – has a bright future?
- a bit too much like Streamtime (but a web app at least)
Tagline: "Let's get to work". It’s the biggest, closest thing to an industry standard, and everyone thinks that it’s the be all and end all… Timesheet stuff works well; and it connects to a whole lot of other things. We might be able to make it work.
- big global player – the standard?
- everything in the world connects to it
- no one ever got fired for picking Harvest
- can't easily turn an estimate into a project
- no per-client hourly rates
- no task categories – we’d need to create a lot of tasks for “estimates”
- can’t assign tasks to people (so having a lot of tasks isn’t ideal)
Tagline: “Easy project management for creative teams". Looks really nice and supports most of what we want.
- looks good, works well
- iPhone and Android apps
- does almost everything we want
- actively being developed, good support
- lacks flexibility (e.g. all invoices show hours)
- too new?
Tagline: “Project management software”. Doesn’t look quite as nice as Trigger, but does have some really nice features.
- simple timesheeting against tasks
- can indicate a “percentage complete” on tasks
- multiple people on a project/task, but one person “pinned”
- there are tumbleweeds in the support blog
- can only timesheet in web app – no widget, or ability to import
- need to create a task before timesheeting
- design/interface quirks
Honourable mention: Although it didn’t make my exclusive shortlist, I thought Breeze was pretty snazzy. The tile-based interface reminded me a bit of a game of memory, and it worked really well using hotkeys – fewer mouse clicks. I got really excited (classic confirmation bias symptom here) and misread “involve clients” as “invoice clients”; sadly, it didn’t do invoices, so it was out.
And the winner…
In the end we chose Harvest. Really? After a year of looking we go for the obvious answer. Hmmmm – maybe there’s a lesson in that! To be honest I’m not completely satisfied with Harvest. You can’t turn an estimate into a project, you have to re-enter the same information manually. It’s not easy to find billable hours that haven’t been billed (something we definitely don’t want to lose!). I could go on. But overall it was the best option available. And it is getting better – their Forecast add-on has potential to really help with resourcing. And if it could help us track potential sales…
Are my expectations too high?