Predict the Future

Pro Tip: Use Onsite Search Data to Predict the Future

Did you know you have the power to predict the future?

It’s true. You can … if you know where to look. You can even manipulate it.

I’m going to let you in on a little secret. There’s a place in your analytics you can use to determine what is likely to happen next – and that’s in your on-site search data.

A lot of the focus on analytics looks at traffic trends and attribution analysis, but in my experience those are useful for showing you what has happened – not what will happen. Yes, you can use predictive analytics on top of your platform to get some better forecasts and predictions, but you also have a treasure trove of information that can give you amazing insight into shifting customer trends and interest already at your disposal.
Read more

Watch Out for Big Data Quicksand

Don’t Get Swallowed by Big Data and Analytics

This past Wednesday I took part in an Adobe chat on Twitter focused on attribution modeling. As usual, members brought a lot of value to the discussion and some great ideas were shared – but one thing in particular stood out to me.

Our need to quantify everything threatens to overwhelm us.

Amidst all the discussion it became clear that even those companies who are data-driven are still struggling to make sense of all the data. This shouldn’t come as a surprise though, as data collection and analysis can easily become a very deep rabbit hole for companies to jump into.

The thing about data is, there’s always more of it. More data points to collect, and more granularity to filter down to. So what we see happening is this: no one ever seems to have enough data. And with that, there’s never enough analysis done. Or at least, that’s the thinking.

But is it maybe possible that we focus too much on the minutiae at the expense of the bigger picture? Even more important, are we really seeing the kinds of returns on data capture and analysis that justify the expense?
Read more

Sad Cyber Monday

Cyber Monday 2016: A Case of Cyber Monday Withdrawal Blues

Today is Cyber Monday. And for the first time since the invention of Cyber Monday … no, since before Cyber Monday even existed, I’m not working for an e-commerce retailer.

And it’s really weird.

I’ve been caught up in a lot of projects lately that have been taking my attention, but today is one of those days where I feel a bit emptier. As the consumer world goes on their mad rush to find the best online deals, e-commerce marketers are busy watching the returns of their long-planned campaigns, hoping their sites don’t go down and that every promotion goes off without a hiccup.

It’s really one of the most exciting days to work in e-commerce … and I’m missing it. Instead of sitting behind a monitor watching sales roll in, cheering for records to be broken and the marketing plan to be a hit, I’m busy working on other things.

They’re fun things, to be sure. Pretty exciting actually (especially a few things that are in the hopper I can’t talk about yet). But they’re not the rush of Cyber Monday.

So all you e-commerce marketers – while you might get a little stressed out today, try to take everything in. It’s the most exciting retail day of the year for you – and probably the one day you’ve been planning around the most over the last few months.

Enjoy it.

Ride the wave.

Celebrate the successes.

Learn from the mistakes.

Try to take a breath.

Or two.

And then remember this. Cyber Monday isn’t even the biggest online shopping day of the year.

That’s not until Green Monday … which for 2016 means you have until December 12th to really start freaking out.

Oh how I miss it.

How’s your Cyber Monday? Or how was it? Any big successes? Colossal failures? Heart attacks? Share below.

A Tip to Make Your Analytics More Actionable

Every month, my team dives into our various analytics tools to put together a bunch of reports to show how the previous month went. This includes marketing channel reports, traffic, conversion rates, social media, app engagement, A/B test findings and a bunch of other data.

Up until recently, that report went into a marketing folder, as well as out via email to the marketing team for everyone to go over on their own to get an idea of how things are working.

This month I made a big change to that approach, and so far it seems to be having great results. So great, in fact, that I want to share them here with you.
Read more

Finding Leaks With Customer Pathing / Flowcharts

As you build out your marketing campaigns and how they integrate into your customer lifecycle flow, one of the first things you’ll do is lose people along the way to achieving your preferred end goal. Face it, you’re going to have fall-off, but not all those people who fall off should be lost. Take a look at your conversion rate. Somewhere in the single digits, right? Yeah, that’s a lot of people then who aren’t converting, and while a good chunk of those people may not have intent to purchase and may actually be people who aren’t worth even trying to sell to, there are a lot who are good potential leads, but they’re getting lost along the way.

Think of it like this: All the people who come to your site go into a bucket. That bucket has a funnel at the bottom and at the end of that funnel is a conversion. Ideally your funnel wouldn’t be funnel-shaped at all, but instead a pipe just as wide as the top, and everyone would flow through quickly and efficiently. Unfortunately, that’s not usually the case. Along the sides of the bucket, as well as within the funnel itself, you’ll find a lot of little holes. These holes are where customers leave you. And once they’re gone, once they’re out of your bucket, they fall off into the nether. These are the leaks in your bucket and funnel.

Your job is to find out where your bucket has leaks, and how you can plug those holes. Or, at the very least, you have to redirect the flow from those holes into a whole new bucket with its own funnel, or pump the lost customers right back up into the bucket and give them another chance to convert (make it through the funnel).

This is where pathing diagrams (aka flowcharts) can come in extremely handy. Explained simply, a pathing diagram is a flowchart of how you want users to go through your experience, from any specific starting point you can define (A),  through any specific “success metric” that’s important to you (B). Your goal is to figure out how the customers go from A to B, mapping out all the ways they might not make it from A to B, and then look for ways to redirect those unsuccessful paths back into a new opportunity to succeed. Each of these unsuccessful paths signifies a leak, and it’s up to you to find a way to capture and recycle as much of what’s leaking as you can.

For example, if a customer lands on your product page, your goal is probably to get them to add to cart, then to go from the cart into checkout and then complete the order.

Of course the customer doesn’t always do that. In many instances the customer is going to just bounce from your site. But you’ve identified that customer as someone potentially interested in what you’re selling, so you try to recapture some of those shopper “leaks” by introducing an email signup so you can reach out to them later with other offers, or you could set up retargeting ads to show that product to the customers as they browse other websites, trying to lure them back. Each of these presents new pathing options for you to design, such as what happens when a customer signs up for an email. Maybe you give them a discount, or maybe you explain your product/service better. Then you try again to get them to convert.

As you draw out your paths and you look at all the potential fall-off points, you’ll likely be overwhelmed with all the possibilities of what a customer could do. My recommendation is to only look at the ones that seem most likely – and you can get a good idea of people’s actual behavior by looking at your web analytics or simply watching some people use your site or app.

The point is this: Until you start really looking at and drawing out all the potential paths customers could take, you’re going to miss some opportunities. If you don’t see those opportunities, you’re never going to address them. It’s easy for us to build out a preferred path for how we want our customers to act, and even to continually optimize to make those paths easier to follow. What you can’t forget though is that not everyone is going to go down your predesigned paths, and unless you want them to wander off, you need to find other ways to help them still arrive at your preferred destination.




Never Underestimate the Power of a Simple Survey (+ a Few Survey Tips)

As digital marketers, we spend a lot of time digging through analytics. And if you’re anything like me, a good portion of that time is spent trying to better understand the customer, their intent, hang-ups on site, or if you’re offering them the right kind of content/product/experience. Of course, analytics can tell you a LOT, especially once you start learning how to die different data sources together to get a more holistic view of the customer.

The one thing they won’t tell you though, is exactly what the customer is thinking. Yes, you can extrapolate and build hypotheses and test for correlations, but in my view the one true way to find out what a customer is really thinking is to just ask them.

Yes, it is true that customers don’t always know what they want, but by asking questions you’ll at least find out what they think they want. And that’s pretty valuable, especially when it comes to determining how customers are making decisions in the conscious part of their brains. But to ask them questions, you need to actually get out there and ASK.

The simplest, and in my opinion, most effective of these is the basic anonymous poll. Put together a question or two and trigger it to display for the people you are trying to understand. Make it as easy to answer as possible, with a limited amount of answer fields so it’s easy for them to make a decision. I also recommend including a free text form for those who don’t fall into the specific categories. You’ll find that some people didn’t quite understand the response options and typed in a variant of one of the existing responses, and you’ll also find additional answers that you never even considered to begin with.

Most importantly though, try to make the answer options as unbiased as you can. Include options that you don’t want to hear (i.e. “I don’t trust your company” or “your prices are too high.”)  Remember, the point of asking questions is to get as honest of feedback as you can so that you have actionable insights you can take back to your team to continue to iterate and improve.

A few tips:

  • For online surveys, try – they have a great system that’s easy to set up, has great insight and analysis tools, and is cost-effective.
  • You can also do surveys/polls directly on Facebook or Twitter, but remember that the only people who will be answering there are the users who actually use those services, which does lead to some potential bias.
  • Use your surveys to build better surveys. As you do a few, you’ll start to understand better how to word questions and responses so they make sense to the customer (as well as avoid leading them to answer with the option you prefer).
  • Your customer support can be a survey. Don’t be afraid to ask your support staff to ask customers questions. Just remember again that you will only be having responses from people who are contacting support, which is not indicative of your overall userbase (unless you have a really bad product and everyone calls support!). Also remember that people are less likely to give honest answers when they are talking to a real person than they are in an anonymous survey.
  • Don’t survey too often. Like anything else that’s not key to the experience, the survey does get in the way of the customer’s goals. Be mindful of this and respect the user.
  • A survey shows you care. Make it clear that you’re asking questions to make things get better. Customers do care that you care about them.

Have you had any positive or negative experiences with surveying? What other ways have you found effective to get direct customer feedback?