1) If it requires > 5 seconds to refresh a page, you should be doing a cron job (or whatever) to generate and then just reading from cached data. If you have any pain in accessing data, you won't make decisions based on data.
2) I like to periodically update my dashboards to a) combat trend blindness and b) include new widgets for projects that have upcoming decisions to be made, like "Should I scale out this new SEO mini-project?" (You'd want data which showed you whether it was working or not.)
3) The point of having eye candy is not to display numbers, it is to make the right decisions. Don't include stuff that isn't actionable. (Exception: need to bamboozle investors? Give them all the data they want and more.)
Disclosure: I run a startup that makes SaaS status boards.
While I agree that having access to _actionable_ metrics is what's really going to be the biggest tangible benefit of these sorts of systems, I've also found that there exists another class of data that, while not actionable, also provides some utility. These are often tagged with the 'vanity metrics' label and conventional wisdom deems them to be useless at best and distracting at worst but, I've found, they can sometimes be useful.
In other words, I may not be able to act (or know how to act) when they change but getting constant feedback from many aspects of my business helps me to notice subtle changes that may be interesting and that sometimes act as a harbinger for larger/more significant changes further down the line.
What these metrics are will entirely depend on your business what sort of person you are. For me, something like changes in my Twitter follower count (a classic vanity metric), when combined with a host of other signals can give me subtle cues about how the business is being perceived, what's causing that, is it a good change etc... none of it is immediately actionable but it all adds up; it's more data points leading a more rounded understanding of how my business works.
It's entirely possible I spend too much time dog-fooding my own product and most reasonable business owners don't have the time to explore these cues, but I do feel like it helps me to feel more connected to the day-to-day running of my own business.
I would be wary of that, given people's incredible facility with fitting a conclusion on top of collections of inconclusive data and then, worse, fitting new data into that conclusion. I have probably done it myself.
Two thumbs up for metrics and startups and startups for metrics, but I hear "delta Twitter followers" and start seeing subtle distinctions teased out holistic evaluation of the arrangement of chicken entrails.
(The difference between Twitter followers and chicken entrails? Chicken entrails have measurable value - people can at least eat them.)
I certainly would not argue that there is explicit value in knowing delta Twitter followers, and wasn't what I was suggesting. If you're trying to retrofit causation to a clutch of assorted metrics you happen to have lying about I totally agree you're going to come up with some misleading conclusions.
What I would say, however, is taken as one - albeit low weighted - input into a system (my brain) that might x-ref against some other indicators of different significance might lead to the conclusion "there's unusual activity afoot". The # of Twitter followers doesn't tell you much about what's happening but might indicate that something is happening.
Our belief is that a small business management Dashboard must be simple. We also recommend, per Jim Collins and every CPA out there, finding your Profit per X or Revenue per X and keeping a keen eye on that number.
Ditto. Whenever a company hides pricing information, I assume the product is incredibly expensive, or that there's some kind of licensing model that involves a phone call and a salesperson. Considering that it's only $10/month, the pricing should really be more prominent.
Agree with the simple. Few of the examples shown above seemed really complicated. It's great to measure your success but does it have to be packed with so many features? I'm not sure.
I would A/B test "Free Trial" vs. "Sign up now for a Free Trial". Also, forcing people to enter a CC number before their Free Trial probably reduces your signups significantly.
Off topic observation on possible front page / ordering glitch: As of 15h50 UTC this submission had vanished from the front page, and wasn't to be found in the second, third or fourth pages. It had "136 points by domino 5 hours ago | 27 comments", which should be enough for the second page.
(It was sitting near the top of the front page; I refreshed and then got curious about it, and had to look down the chronological list to find it again).
This is a phenomenal post. I completely agree that it improves things - we recently got a massive TV in the office at Distilled to show off some key metrics using Geckoboard: http://yfrog.com/h3uhzrj
Any recommendations for an open source dashboard? Like what Piwik is for Google Analytics, but only in dashboard-land like the ones in the original post?
(Yeah, money is tight...)
Edit: Except the ones mentioned at the bottom of the post, of course.
Aw man this is awesome, I've been secretly longing for a visual dashboard for months now, but for one reason or another I never got around to having one.
Never realised there were so many out-of-the-box solutions already, this makes getting a good dashboard a weekend project! Sounds like a challenge ...
You really can't talk about dashboards without talking about the Open Source Business Intelligence Reporting Suite that http://jaspersoft.com (http://jasperforge.com) offers.
If not a dashboard, then it's periodic reports... Or reports pullable on demand.
There's got to be a way to measure things and plan for the future, and reports/graphs/etc help a lot with that.
I think reports are necessary, but a dashboard really helps everyone see at a glance how things are going.
If your up-time graph looks like a heartbeat monitor, everyone knows what that means.
Likewise for active customers being a downhill slope when viewed weekly.
Most of the time, those graphs are just going to give you happy thoughts, but sudden spikes or dips, or worrisome things like those above... Well, you can SEE them and they mean more, especially to non-technical types.
Yes, absolutely every startup I've talked to creates their own dashboard or dumps some results to excel periodically. And I've talked to a lot of them, as my company is addressing this issue. (chart.io YCS10)
Dashboards are awesome. That said I think they can easily become a glitter project that detracts from more critical work. Shell scripts, cron+email, or a bare-bones web page can all be cheaper alternatives and are quite useful.
Disclosure: I built InfoCaptor - Dashboard software
most small businesses still use Excel and Access for their data. They consolidate their excel files from various sources like quickbooks, other spreadsheets. But if you are just starting, the only thing that matters is to measure the user count and number of sales.
It is only when you start spending for marketing, outsourcing etc you need to start keeping tab on your "profit margin","revenue" etc
1) If it requires > 5 seconds to refresh a page, you should be doing a cron job (or whatever) to generate and then just reading from cached data. If you have any pain in accessing data, you won't make decisions based on data.
2) I like to periodically update my dashboards to a) combat trend blindness and b) include new widgets for projects that have upcoming decisions to be made, like "Should I scale out this new SEO mini-project?" (You'd want data which showed you whether it was working or not.)
3) The point of having eye candy is not to display numbers, it is to make the right decisions. Don't include stuff that isn't actionable. (Exception: need to bamboozle investors? Give them all the data they want and more.)