Gelato was the word I meant

I spent most of the last week on holiday in Italy. But since the holiday was built around a speaking gig in Italy at the Be Wizard Digital Marketing conference I still spent a couple of days talking analytics and digital. A couple of days I thoroughly enjoyed. The conference closed with a Q&A for a small group of speakers and while I got a few real analytics questions it felt more like a meet and greet – with plenty of puff-ball questions like “what word would use to describe the conference?” A question I failed miserably with the very pathetic answer “fun”.

I guess that’s why it’s better to ask me analytics questions.

The word I probably should have chosen is “gelato”.

And not just because I hogged down my usual totally ridiculous amount of fragola, melone, cioccolato, and pesca – scoop by scoop from Rimini to Venice.

Gelato because I had a series of rich conversations with Mat Sweezey from Salesforce (nee Pardot) who gave a terrific presentation on authenticity and what it means in this new digital marketing world. It’s easy to forget how dramatically digital has changed marketing and miss some of the really important lessons from those changes. Mat also showed me a presentation on agile that blends beautifully with the digital transformation story I’ve been trying to tell in the last six months. It’s a terrific deck with some slides that explain why test&learn and agile methods work so much better than traditional methods. It’s a presentation with the signal virtue of taking very difficult concepts and making them not just clear but compelling. That’s hard to do well.

Gelato because I also talked with and enjoyed a great presentation from Chris Anderson of Cornell. Chris led a two-hour workshop in the revenue management track (which happens to be a kind of side interest of mine). His presentation focused on the impact of social media content on sites like TripAdvisor on room pricing strategies. He’s done several compelling research projects with OTAs (Online Travel Agents) looking at the influence of social media content on buying decisions. His research has looked at key variables that drive influence (number of reviews and rating), how sensitive demand is to those factors, and how that sensitivity plays out by hotel class (turns out that the riskier the lodging decision the more impactful social reviews are). He’s also looked at review response strategies on TripAdvisor and has some compelling research showing how review response can significantly improve ratings outcomes but how it’s also possible to over-respond. Respond to everything, and you actually do worse than if you respond to nothing.

That’s a fascinating finding and very much in keeping with Mat’s arguments around authenticity. If you make responding to every social media post a corporate policy, what you say is necessarily going to sound forced and artificial.

That’s why it doesn’t work.

If you’re in the hospitality industry, you should see this presentation. In fact, there are lessons here for any company interested in the impact of reviews and social content and interested in taking a more strategic view of social outreach and branding. I think Chris’ data suggest significant and largely unexplored opportunities for both better revenue management decisions around OTA pricing and better strategies around the review ask.

Gelato because there was one question I didn’t get to answer that I wanted to (and somehow no matter how much gelato I consume I always want a little more).

Since I had to have translations of the panel questions at the end, I didn’t always get a chance to respond. Sometimes the discussion had moved on by the time I understood the question! And one of the questions – how can companies compete with publishers when it comes to content creation – seemed to me deeply related to both Mat and Chris’ presentations.

Here’s the question as I remember it:

If you’re a manufacturer or a hotel chain or a retailer, all you ever hear in digital marketing is how content is king. But you’re not a content company. So how do you compete?

The old-fashioned way is to hire an agency to write some content for you. That’s not going to work. You won’t have enough content, you’ll have to pay a lot for it, and it won’t be any good. To Mat’s point around authenticity, you’re not going to fool people. You’re not going to convince them that your content isn’t corporate, mass-produced, ad agency hack-work. Because it is and because people aren’t stupid. Building a personalization strategy to make bad content more relevant isn’t going to help much either. That’s why you don’t make it a corporate policy to reply to every review and why you don’t write replies from a central team of ad writers.

Stop trying to play by the old rules.

Make sure your customer relations, desk folks, and managers understand how to build relationships with social media and give them the tools to do it. If you want authentic content, find your evangelists. People who actually make, design, support or use your products. Give them a forum. A real one. And turn them loose. Find ways to encourage them. Find ways to magnify their voice. But turn them loose.

You can’t have it both ways. You can’t be authentic while you try to wrap every message in a Madison Avenue gift wrapping bought from the clever folks at your ad agency. Check out Mat’s presentation (he’s a Slideshare phenom). Think about the implications of unlimited content and the ways we filter. Process the implications. The world has changed and the worst strategy in the world is to keep doing things the old way.

So gelato because the Be Wizard conference, like Italy in general, was rich, sweet, cool and left me wanting to hear (and say) a bit more!

And speaking of conferences, we’re not that far away from my second European holiday with analytics baked in – The Digital Analytics Hub in London (early June). I’ve been to DA Hub several years running now – ever since two old friends of mine started it. It’s an all conversational conference modeled on X Change and it’s always one of the highlights of my year. In addition to facilitating a couple conversations, I’m also going to be leading a very deep-dive workshop into digital forecasting. I plan to walk through forecasting from the simplest sort of forecast (everything will stay the same) to increasingly advanced techniques that rely, first on averages and smoothing, and then to models. If you’re thinking about forecasting, I really think this workshop will be worth the whole conference (and the Hub is always great anyway)…

If you’ve got a chance to be in London in early June, don’t miss the Hub.

Big Data Forecasting

Forecasting is a foundational activity in analytics and is a fundamental part of everyone’s personal mental calculus. At the simplest level, we live and work constantly using the most basic forecasting assumption – that everything will stay the same. And even though people will throw around aphorisms of the “one constant is change” sort, the assumption that things will stay largely the same is far more often true. The keyword in that sentence, though, is “largely”. Because if things mostly do stay the same, they almost never stay exactly the same. Hence the art and science of forecasting lies in figuring out what will change.

Slide 1 ForecastingBigData
Click here for the 15 minute Video Presentation on Forecasting & Big Data

There are two macro approaches to forecasting: trending and modelling. With trending, we forecast future measurements by projecting trends of past measurements. And because so many trends have significant variation and cyclical behaviors (seasonal, time-of-day, business, geological), trending techniques often incorporate smoothing.

Though trending can often create very reliable forecasts, particularly when smoothed to reduce variation and cycles, there’s one thing it doesn’t do well – it doesn’t handle significant changes to the system dynamics.

When things change, trends can be broken (or accelerated). When you have significant change (or the likelihood of significant change) in a system, then modelling is often a better and more reliable technique for forecasting. Modelling a system is designed to capture an understanding of the true system dynamics.

Suppose our sales have declined for the past 14 months. In a trend, the expectation will be that sales will decline in the 15 month. But if we decide to cut our prices or dramatically increase our marketing budget, that trend may not continue. A model could capture the impact of price or marketing on sales and potentially generate a much better prediction when one of the key system drivers is changed.

This weekend, I added a third video to my series on big data – discussion of the changes to forecasting methodology when using big data.

[I’ve been working this year to build a legitimate YouTube channel on digital analytics. I love doing the videos (webinar’s really since they are just slide-shows with a voice-over), but they are a lot of work. I think they add something that’s different from either a blog or a Powerpoint and I’m definitely hoping to keep knocking them out. So far, I have three video series’ going: one on measuring the digital world, one on digital transformation in the enterprise, and one on big data.]

The new video is a redux of a couple recent speaking gigs – one on big data and predictive analytics and one on big data and forecasting. The video focuses more on the forecasting side of things and it explains how big data concepts impact forecasting – particularly from a modelling perspective.

Like each of my big data videos, it begins with a discussion of what big data is. If you’ve watched (or watch) either of the first two videos in the series (Big Data Beyond the Hype or Big Data and SQL), you don’t need to watch me reprise my definition of big data in the first half of Big Data and Forecasting. Just skip the first eight minutes. If you haven’t, I’d actually encourage you to check out one of those videos first as they provide a deeper dive into the definition of big data and why getting the right definition matters.

In the second half of the video, I walk through how “real” big data impacts forecasting and predictive problems. The video lays out three common big data forecasting scenarios: integrating textual data into prediction and forecasting systems, building forecasts at the individual level and then aggregating those predictions, and pattern-matching IoT and similar types of data sources as a prelude to analysis.

Each of these is interesting in its own right, though I think only the middle case truly adds anything to the discipline of forecasting. Text and IoT type analytics are genuine big data problems that involve significant pattern-matching and that challenge traditional IT and statistical paradigms. But neither really generate new forecasting techniques.

However, building forecasts from individual patterns is a fairly fundamental change in the way forecasts get built. Instead of applying smoothing techniques for building models against aggregated data, big data approaches use individual patterns to generate a forecast for each record (customer/account/etc.). These forecasts can then be added up (or treated probabilistically) to generate macro-forecasts or forecasting ranges.

If you’ve got an interest in big data and forecasting problems, give it a listen. The full video is about 16 minutes split into two pretty equal halves (big data definition, big data forecasting).