Posts Tagged ‘data’

10 Tips for small business: surveying your customers

Tuesday, May 3rd, 2011

One of the most important things a business can do is to invest the time to understand its customers. This can be accomplished in several ways: by collecting and analyzing business data, by observing customer behavior, or by studying comparable companies. Or, you could always just take the simple approach and ask them. This simple, straightforward method can be done in a few different ways: by speaking to your customers one-on-one as they visit you or you visit them; by convening a small focus group of your customers and hiring an experienced moderator; or by creating an online survey and asking your customers to participate.

All three approaches can give you excellent results at varying cost and reliability, although the first is not particularly scientific. At crowdSPRING we have been very successful with targeted surveys designed to answer specific questions, or to gather detailed information; on separate occasions, we have surveyed our buyers to better understand their demographics; to get a clearer picture of their satisfaction; and to specifically understand why some projects fail. We have also surveyed separate groups of Creatives on the site to get a clear picture of their professional backgrounds; to understand what new tools they would like us to develop; and to clarify exactly what motivates them to participate.

We approach these surveys scientifically and always recruit a large enough group of participants to provide a statistically meaningful sample (you can read more about this here). We have learned a great deal through reading and understanding best practices, but  also (and of equal importance) through trial and error about how to run a successful survey. I’ve put together some tips and tools that will be useful to you as you plan the process of surveying your own customers and clients. Good luck!

1. Define a goal.
The most important thing you can do when planning a survey is to decide what exactly you want to know. Is this about customer satisfaction? Behavior? Demographics (or “firmographics”)? Take the time to determine what you wish to understand and plan your survey accordingly – your goals should, in large part, drive the design and execution of your survey.

2. Recruit the right participants.
If you are fortunate enough to have a large user base, take the time to segment properly. For instance, if the goal of the survey is to understand your customer’s attitudes towards medicare you might want to only survey those over 55 years of age. Alternately if you want to know more about your customers tastes in indy rock, you may want to target the under 30 set.

3. Incentivize properly.
You say you want participation? Toms of responses? Piles of data to analyze? Well then you’ll need to offer something of value in return. Professional market researchers will tell you that you will receive a significantly higher response rate if you offer a prizes (or compensation of some sort) for participation; even a token award or modest raffle will increase response rates meaningfully.

4. Craft your questions well.
Remember that the quality of the responses you receive is in large part a measure of the quality of the questions you ask. First thing first: keep em short and easy to understand; participants will have little patience with long or confusing questions. Next, start with the easy questions, then move on to the more complex – this will serve to draw the respondent into your survey and lead to higher completion rates. Use a variety of question types: multiple choice checkboxes, likert scale, yes/no; by mixing these up you will hold the participants interest and lead to greater engagement and more accurate responses. Finally take care that your questions are not leading or biased; remember you want their answers, not just the ones you’d like them to give.

5. Keep it short, make it fun.
A good survey should take less than 10 minutes for the respondent to complete. Five is even better. For the best results, make sure that your survey is focused, goal oriented (see #1 above), and above all, BRIEF!

(more…)

Not Everything That Can Be Counted Counts

Monday, October 11th, 2010

It’s tempting to count everything that can be counted. The current obsession with social media ROI is just one example – of many – showing how numbers influence our behavior.

I had the opportunity to again think about this issue last week when I was on a panel at a Startup Bootcamp event in Boston. We discussed, among other things, data analysis and how some startups regularly look at the data to learn insights and help plan strategy.

I expressed my view: trying to count everything can be wasteful, distracting and can lead to decision paralysis. Startups and small businesses that become obsessed with metrics often lose their way.

Not all metrics are important. Albert Einstein famously said:

Not everything that can be counted counts, and not everything that counts can be counted.

Wise words – most startups and small businesses should take those words to heart.

Data is important, but you should let your assumptions and theories drive your analysis of the data, not the other way around. (for some tips on working with data, I recommend you read 5 thoughts for startups and small business: numbers count.

We tend to rely very heavily on metrics crowdSPRING andcan easily become distracted if we don’t smartly pick and choose the metrics that influence our decisions. Sometimes, we make the right decisions and focus on the right metrics. Other times, we make the wrong decisions and lose focus, paying attention to metrics that turn out not to be as relevant as we thought. (Last week, I suggested four questions you should ask when making decisions based on metrics and statistics).

Given the wide availability of good software and plenty of data (from your internal and from many external sources), it’s pretty easy for startups and small businesses to measure pretty much anything.

One of the lessons we’ve learned from our successes and failures: we are more likely to succeed when we spend a greater portion of our efforts discussing and debating what should be counted – and a smaller portion of our effort counting.

Numbers are good – but everything does not need to be counted.

Do you agree?

Small business and startup tips: track your own customer service data

Monday, July 12th, 2010

It’s been over a year since I last updated the community on our efforts to build a world-class customer service organization. We are committed to providing great support and doing it in a timely, polite, and efficient manner. Since we launched in May of 2008 we have provided customer service via our online Help Center, our help desk software via the Contact Us form available anywhere on the site (word to our friends at Zendesk!), through Private Messaging, and more recently over the phone (albeit in a limited fashion). To date we have received over 47,000 requests through the site along with thousands more PMs, emails, and calls! Sometimes the folks are confused, sometimes friendly, occasionally really pissed, but we’re happy to answer their questions, help solve their problems, (hopefully) assuage their anger, and, certainly, play along with their fun!

Small businesses and startups have much to gain by providing great support and building lasting relationships with their customers. Loyalty, word of mouth, and brand value are all increased when a business gives its customers information, attention, and answers. In order to constantly improve our own responsiveness, we review and analyze our own support data quarterly and consider the implications on staffing, tools, features, page design, support channels, and quality of service. We learn a great deal through these exercises and believe that all businesses should perform similar audits periodically on their own data. So, without further ado, here’s a selected analysis of the customer service data we look at with a focus on the “when,” the “what,” the “how we doin?” and the volume of requests (AKA “OMG, can we handle this much?”)

1. When.

We always look at the requests that come in relative to our own work hours and consider how we allocate staffing to make sure that customers are answered in a reasonable amount of time. The charts below show a breakdown by day of the week and the time of day.

2. What.

We also look hard at exactly what people are contacting us about and consider this data when we make decisions on improvements, new features, community engagement, and site policies. In response to this information, last mOnth we introduced a new Contact Us form/feature which makes it easier to get in touch and also serves up additional Help content and answers critical FAQs inline. At the same time, we realized that, by changing up the categories of questions (via the pulldown on the Contact Us form) we could improve our understanding of our user’s issues and deliver better, stronger, and more convenient content to get their questions answered and their problems solved.

(more…)

Saying “thank you” – why startups and small businesses should listen to mom.

Monday, June 14th, 2010

In these recessionary times, all of us who operate startups and small businesses struggle to meet our bottom line, pay our monthly bills, retain our great employees, and grow our businesses in spite of the challenges presented. What to do, what to do? Well, one time-proven marketing strategy that increases customer lifetime value, cements loyalty, and drives word of mouth, is simply doing what your mom tried to teach you to do always: say “thank you.” It is a truism that it costs more to acquire a new customer than to retain an existing one, and this simple strategy goes a long way towards this.

According to a survey of small businesses by American Express, only one third of businesses gave year-end bonuses or gifts to their employees last year, down more than 10% over the prior two years. Recession, ugh. Yet many small businesses continue to build customer loyalty and increase sales by giving gifts to their customers. The gifts can be anything from swag like tee shirts and coffee mugs, to discounted services or merchandise. Studies show that this works, and the credit card industry is a good example of this. For instance, in a study by the consulting firm Maritz, credit card holders who participated in loyalty programs increased their number of transactions by 53 percent and the value of  those transactions by 51 percent!

How can your business say thank you to your customers? Experiment. That’s right – try lots of different things to see what works best. Try a loyalty program; my local pizza place does it – every time I go in they stamp my card. Ten stamps and I get a free pizza. Yum. Or try an A/B test: offer one group of your customers a discount on their next purchase, and offer the another group a tee shirt. Then track the results – which group came back more often and spent more money? Or try a different experiment; offer some customers a gift card for a successful referral of a new customer and an equal number a small cash bonus. Which group responds best? The answer to this question could define your strategy for your next round of “thank-yous.”

Here are a few tips for thinking through your approach to gifting your customers:

  • Segment. Look hard at your customer base and consider what type of incentive might be valuable to what type of customer.
  • Communicate. Ask for feedback when you give a gift or say your thank you. If you ask them what they like and why, many of them will actually answer you.
  • Analyze. Don’t just send out gifts without paying attention to the results. Track the data and be ready to dispense with the gifting that doesn’t work and beef up that which does.
  • Get buy-in. Make sure that all of your key departments or employees are involved, from the front-line customer service folk to the marketing and sales departments.
  • Provide value. If you are looking for insight into hat makes your customers tick and what increases their lifetime value, only they can tell you. And they will only tell you if the thank you provided is of value to them.
  • ROI. Make sure that the payouts you are offering are generating more than the cost. It would be silly to offer every customer a $100 gift card if the increased value of that customer were only $50. Be careful here and make sure you end up profiting on these efforts.

So, dangle the carrots and see if they nibble. And, like mom said, “Be polite. Or I’ll smack you.”

Photo credit: Michael Newman

5 thoughts for startups and small business: numbers count

Monday, March 8th, 2010

Photo: tifotter

Einstein famously said, “Not everything that can be counted counts, and not everything that counts can be counted” while Count Von Count (pictured above) said “That’s TWO! Two worried frogs! Ha ha ha ha!” Einstein was right in so many ways, but I believe that in the context of small businesses and startups, the Count was the one who truly grasped the full import of collecting and leveraging business intelligence data. In today’s world of software and data, it is so easy to count everything which might be of value, that there is no reason not to. The trick is in how you leverage that information, and how you design access to it. In other words: count everything you can, but take the time to determine how to prioritize that which you do count, and what to pay attention to when you look at that data.

As I have discussed here before, I am a proponent of a goals-driven framework for business strategy and tactics. I believe that strategy should be devised to help meet meaningful goals and that tactics should be implemented to support the strategy. The key ingredient in goal-setting is a defined metric, and the ability to measure progress towards that goal is via (duh) the data that must be tracked  and gathered. In the absence of meaningful information, decisions can easily be made for the wrong reasons and priorities can be improperly set; managers need to make rational decisions and let those decisions be driven by data first, and intuition second. Many owners and managers of small businesses find that they make their decisions by instinct, that they  “fly by the seat of their pants,” often in the absence of the meaningful data which should help to drive those decisions.

Here are 5 tips for gathering and leveraging data to drive a rational approach to defining and meeting goals and formulating and executing strategy.

1. Debate and strategize. First determine what it is you are trying to accomplish and how you intend to get there. it could be something as simple as increasing revenue;  improving the monthly traffic to your site; or gaining a better understanding of your customers to improve satisfaction. No matter what it is, first debate with your team exactly what the important goals are, and make sure they are measurable based on the data you have the ability to collect. Set a strategy to accomplish the goal: more revenue? Maybe the strategy is to increase lifetime value of your customers, or maybe it is to increase per-order revenue. More traffic? maybe the strategy is to improve word-of0mouth. Better customer satisfaction? Maybe the strategy is to reduce wait times for customer service requests. In any case, measure, measure, measure.

2. Collect and organize. This is a process which can be as difficult or as easy as you make it. At crowdSPRING we have a database that is fairly easy to query and a process for creating new queries quickly. This allows us to set up simple tools to pull specific types of data using date ranges. For instance, we pull a weekly report on all new projects posted, which includes some immediately useful data (how many projects were posted last week, what was the value of those, which buyers posted them) and some data which might come in useful later (time of day the project posted, IP address of the buyer, date/time of buyer’s registration on the site) . This data is aggregated in a spreadsheet, which allows a quick update every week.

3. Sort and prioritize. The data collected must be sorted to be useful; for instance the New Projects report I mention is default sorted by date and time, because it is our priority to understand projects and revenue on a daily and weekly basis. If it were our priority to understand average award in projects, we would default to that.

4. Display and visualize. Different people look at data differently. I tend to consume mine best in the form of a chart or a graph, while Ross prefers to eat his data raw, just the numbers in front of him. Excel and other spreadsheet programs give you simple and powerful tools for setting up charts, graphs, and tables, and these can be leveraged in ways which make analysis simple and fast, which leads to the next point:

5. Analyze and theorize. Data in a vacuum is meaningless; we use the data we collect to measure our progress against goals, and to determine the efficacy of our strategy and tactics. Some analysis is simple (how many projects posted last week),  but it can get very complex. We typically theorize the impact a new tactic will have on our progress towards a goal and we then measure our progress from the date of that implementation. This analysis helps us to determine whether to continue an effort, abandon it, or modify it. When a small business is constrained by its own capacity (as so many are) it is critical to know when to move on.

Collect your data – all of it – but use it to serve a greater master: your goals and your strategy.

Small business tips and tricks: reconciliation hell (or 12,716 ways to spend your weekend)

Monday, February 1st, 2010

Warning to readers: the topic of this week’s blog post is arcane, technical, and probably boring to roughly 95% of our audience. I hope that you “5-percenters” get something out of this and that the other 95% of you appreciate that what I am about to describe is how many entrepreneurs spend their Sunday afternoons in late January.

12,716 transactions in 2009. Separate transactions. Individual transactions. This is what we did in 2009 and now our accountants insist that each and every one of these reconcile our own data against our merchant processor data. OMG!

Why is this ridiculous exercise necessary? Well, they tell us that if we are ever audited non-reconciling payments will prove problematic. Problematic? Hello! Problematic is finding a way to do this that won’t drive me over the edge…

Alright, I exaggerate. There are ways to do this with relative ease. The key word being “relative.” Here’s what I did to ease my load and (fairly) quickly identify the problem transactions. Our merchant processor is PayPal. PayPal is great in many ways, but surprisingly awkward to use in many others. Every month I can download a report from their site which gives me a list of all of the transactions generated on our account for that month. Problem is that they do not make it easy to customize these reports, so you pretty much have to live with what they make available. OK – I can deal with that.

So I download roughly 1,000 transactions per month. Checking these manually, one by one, would literally take hundreds of hours, so I John helped me to create a new database report: “All transactions” is a simple list, which I can import into excel and that includes information such as the project name for each transaction along with date, time, amount, and (most importantly) PayPal transaction number. I simply import these into a spreadsheet one month at a time and I’m almost ready to go.

Next I took the PayPal transaction reports and did the same thing: imported them into another spreadsheet, month by month. Now begins the fun. If you have never used the Excel VLookup function, you may want to learn a bit about it. This powerful tool allows you to search across multiple spreadsheets to find matching information. By telling the “PayPal Transactions” spreadsheet to look for the matching transaction numbers in the “All Transactions” spreadsheet i can make excel return the name of the project associated with the PayPal transaction number. Simply copy the formula down the list and, voila! The project names magcally appear on the PayPal spreadsheet. Any transactions that don’t register a match display as “#N/A.” (BTW, what is up with Excel and these bizarre values that it returns? Why can’t it just say “error” or “no value?” Whatever, I don’t have time to teach Microsoft about user experience.)

Now I can simply sort each month’s data by the VLookup column forcing the #N/As to the top. Fortunately for me, there are probably only 5 or 10 for each month and this week I will attack those to figure out what they were. Here’s a link to a great post with step-by-step instructions for using VLookup. Take a look and see if this can help you to match up your own data, whether it is financial data, customer data, or just your collection of DVDs.

And I do hope that our accountants appreciate the effort. Dara? Are you reading this?

Dear Buyer (redux), or 4 tips for posting and managing your crowdSPRING project

Monday, November 16th, 2009

Seven months ago I wrote a post and shared some data which I hoped would help crowdSPRING Buyers to find success and happiness with their projects on the site (well maybe not happiness, but you get my drift). Well, since then we have had more than 5,800 new Buyers registered, had over 336,000 more entries to projects, seen another 2,900+ projects completed, and paid out more than $1,653,000 in awards to Creatives!

We’ve learned a lot more about how things work around here and had some interesting insights about what works and what doesn’t. So, dear Buyer, please take a look at the information below and see if this will help you when posting your next project.

Remember how a few sentences ago I mentioned insights ? Well chief among them is what I am calling the “Rule of Four.” The Rule of Four is, simply enough, four basic things every Buyer should consider when posting and managing a project on crowdSPRING.  Here they are… ready? OK:

  1. Choose a great title for your project, the “sexier” the better. Remember that the title is the first thing Creatives will see when we send out notifications and when they browse the current projects on the site. Given a choice between a project named “Redesign of a Logo” and another one named “Hi! My name is FastLaunch Logo. Please design me. I don’t exist yet in visual form, but I’d like to” which would you go take a look at first? ‘Nuf said about that…
  2. Be generous with your award offer. Remember you’re competing with all of the other open projects on the site, and all else being equal, the award may end up being the thing that brings ‘em in. Here’s some raw data on the topic: logo design projects with awards under $300 average around 70 entries while those with offers over $300 averaged 193 entries! wow. Same with web design: projects with awards under $700 average around 39 entries while those with offers above averaged 63. Can you see a pattern emerging? How about something else: want more Creatives to participate? Those same logo projects that offered under $300? On average 26 Creatives participated. The projects with awards over $300? 57 participants per project. Yep, more dollars = more participation = more entries = greater choice for you.
  3. Write a strong creative brief. Ah yes, the brief. First thing to remember here: it shouldn’t be brief. There is no place for brevity in a crowdSPRING creative brief. The more information you can share the better. The more detail you supply the better. Examples of work you like? Check. Uploaded files to provide even more options for participants? Check. Links to your current site or other information about your business? Check. The point here is that no one can design in a vacuum. So fill the vacuum up with information that they can use.
  4. Be engaged. I’m not talking about giving your Creative a diamond ring. I’m talking about your level of engagement with your project. Do you visit every day? Do you score and comment on every entry (or at least most of them)? This is probably the most critical element in managing your project and you’d be surprised how many Buyers fall down on this job and how, as a result, their projects suffer. As a matter of fact, we see a VERY strong correlation between feedback and entries. It’s simple: more comments = more entries. Here’s some numbers for you to digest: Buyers who left fewer than 10 comments in their project averaged 52 entries to choose from. Buyers who left more than 10 comments? Ready for this? 212 entries per project. 212 entries from which you can pick the one, as long as you leave feedback! So do it. Stop reading this and go leave some comments, for crying out loud.
So, that’s it. Pretty simple right? Think up a great title, offer as much as your budget can stomach, write a high-quality, detailed brief, and stay active and engaged in your project. Follow these four simple rules and you will be in like Flynn! So do it. Stop reading this and go leave some comments, for crying out loud.

Friday Fun Facts – Customer Service Requests

Friday, March 27th, 2009

As a small business startup we spend a great deal of time looking at and analyzing data to help us drive strategy, make adjustments, and better understand our customers. We try hard to not be robotic in our response to numbers, but rather to let them inform our decisions and our direction. Here is the latest in a serial discussion of some of the data we look at internally…

Just back from South by Southwest a few days ago and boy am I excited to get back to some heart-pounding data! This week we’ll take a look at customer service requests. You know, these are the little notes you send to us when you click on the Contact Us link at the bottom of any page. With the arrival of Jaclyn (DBA ‘Snow White’) last week, Jerome thought it would be great to get a better handle on which days of the week are busiest. This was pretty easy to figure out. I created a table which contained the number of support requests we received per day over a three month period. From there Excel helped me to determine how many of the tickets were created on each day of the week. Take a look at the results:

Now Jerome is not a guy who is easily satisfied. Now that we had figured out which were the busiest days of the week for him, he wanted to know which were the busiest hours. The guy is relentless. This was a bit trickier. In order to do this I needed to look at the time stamp on every ticket that came in. To pull this data for a three month period was just too much (I do have other responsibilities, Jerome) so I cheated and gathered the data for the last two weeks only, knowing that this would be a large enough sample to give the French Guy the info he was after. Again, I created a table with columns for time stamps, ticket category, ticket number, and description. Turns out that almost 60% of all requests come in between 1PM and 10PM GMT (7AM and 4PM Chicago time – business hours, how convenient):

Alright, Jerome – let’s make it interesting. What are the busiest hours for Bugs reports? Here ya go – more than 60% of these between the hours of 3PM and 8PM GMT:

What about reports of IP violations? Easy. 60% come in between 12PM and 10PM GMT:

And finally, when do our users think up great ideas for ways to make crowdSPRING better? Here’s the distribution for Suggestions, with around 65% sent to us between 1PM and 11PM GMT:

I hope the French Guy is happy and I hope that Jaclyn can be available to help with the coverage so we can maintain our reputation for speedy response times. Jerome? Happy now?