BIG DATA in events

Ever increasing scores of companies are developing ways of – or are already longtime active in – offering their products and services online.  Companies are investing growing amounts of time and resources to online marketing strategies and campaigns to leverage customer loyalty and engagement in order to keep ahead of the competition.  Organizers of meetings and events are reluctantly slow adapting to the same practices.

One of the most important steps towards ultimate customer loyalty is in-depth, metrics based customer segmentation.  Going online with your company implies a larger, more international and heterogeneous customer base and thus more data.  But even for your real life event offer, you will need to know your customer better in the future, just to retain his/her attention and desire to participate.

Facing an explosion of available and collectible data on each individual customer, achieving solid segmentation becomes an even more challenging task.  So the question emerges, is big data worthwhile ?

Bit size data segments

To keep big data manageable, it’s considered to be good practice to breakdown the data.  Most client data collected online can be grouped in these categories :

• contextual data : interests (identified and profiled), web pages visited, visited sponsors/exhibitors, …
• attitudinal data : number of interactions, number of clicked ads or banners, number of visited pages,…
• behavioral data : number of recurrent visits to your site, average visit duration, type of product bought,…
• demographic data : age, sex, address, Facebook or Twitter id,…
• loyalty data : number of complaints, number of recurring event visits, engagement score,…
• value data : average amount spent on products, score for expected profitability,…
• technical data : used operating system, used browser, Flash version,…
• referral data : referral page, used search words,…

Breaking down the data into these categories has proven very effective to setting and achieving marketing goals.  Academic research has revealed that data segmentation supports the most important online marketing goals, such as increasing up- and cross-selling numbers, attracting new clients and improving churn rates.  The categories of attitudinal and behavioral data are well suited for supporting almost every goal, whereas demographic data is less effective, because of privacy restrictions prohibiting massive online collection of this type of data.  On top of that, although it’s usually the type of data that’s used when segmenting for events, it’s totally irrelevant towards target groups’ interests.

If after reading this part, you’re still wondering what big data is and how it distinguishes from regular data, then read my previous blog post (in Dutch).

Big data challenges

Companies are feeling overwhelmed with massive amounts of data and information.  So how do you manage that ?  Before this explosion occurred, companies were faced with less amounts of data and mostly structured, primarily disclosed in such applications as ERP (Entreprise Resource Planning) and CRM (Customer Relationship Management).  Today, together with the breakthrough of internet based platforms, not only the size of databases has increased tremendously, but also the complexity of the information.  Data can nowadays be stored in structured or unstructured ways, in video, audio and pictures, not just in text.  Managing this and turning it into valuable insights, requires new work methods.

Big data and traditional data not only vary in volume, structure and transaction speed, but also require new tools and technologies.  Today many data management techniques are used in big data that have a common background from the techniques used in traditional data : classification, clustering, regression, time series analysis and A/B testing, to name a few.  But Big Data demands new techniques that empower companies to store, process and analyse vast amounts of unstructured data originating from different sources in various types and sizes.  Big data also requires new and improved visualization methods and tools.  Ways to display all that collected data that are comprehensive for humans, so they can actually use it to make business decisions.

Is big data a joy or a burden ?

It’s a joy if you find the appropriate way to derive knowledge from the data.  Segmenting client data plays an important role in this game.  It will substantially contribute to making better and semi-automated decisions on marketing strategies and tactics to increase the customer engagement.

Start planning for big data today if you’re in the event industry.  There’s no way it will not impact your business in the future.  Big data is directly linked to the way your future customers behave today.  Look for my next blog post in which I will try to depict your John Doe customer of 2015.

Don’t take BIG DATA lightly or as a pure IT-problem and get expert help.

 

Wat is BIG DATA ?

Je kan er niet omheen, tenzij je een tijdje weg was naar Mars.  Iedereen in het bedrijfsleven heeft het erover in de één of andere strategische meeting.  Elke marketing-guru praat erover.  In alle IT-hokken gonst het er van.  BIG DATA, BIG DATA.  Je moét er wel mee bezig zijn of je wordt aangekeken als een sukkeltje.

Allemaal goed en wel, en waarschijnlijk zal het wel erg belangrijk zijn, maar wat is dat nu precies ?  Waarom is het zo belangrijk en hoe pak ik die BIG DATA dan aan ?  Ook in de eventwereld veel gehoorde vragen, bij kleine én grote bedrijven.

Eerste vraag is echter, waarom heeft iedereen het erover ?  Wel, omdat het écht wel actueel is.  De digitalisering van de wereld, het feit dat die digitalisering z’n weg heeft gevonden naar je broekzak (dank je wel Apple en Android) en het drastisch toegenomen gebruiksgemak van de tools die toegang geven tot het digitale, maken dat er vandaag tonnen data en informatie beschikbaar zijn over elk individu.  Bovendien heeft die digitalisering ervoor gezorgd dat het aantal individuen dat je als bedrijf tot je doelgroep mag of kan rekenen, exponentieel gestegen is.  Digitale business assets worden namelijk niet of veel minder beperkt door een geografisch gebied.

Moet je dáárom met BIG DATA bezig zijn ?  Neen, niet noodzakelijk.  Maar tegelijkertijd heeft ieder individu door de informatie-overstroom – terecht – steeds strengere eisen gesteld aan de relevantie van informatie, alvorens er de nodige aandacht aan te besteden, laat staan de gewenste en bedoelde actie aan te koppelen.  Zeker wanneer hij/zij de informatie niet zelf zoekt maar aangereikt krijgt.

Dáárom moet je er mee bezig zijn.  De drempel om de aandacht, die je zoekt, te krijgen is namelijk fors verhoogd.  Je klant écht goed kennen en z’n interesse weten te behouden door relevante inhoud aan te bieden, op allerlei manieren, daar draait het in de toekomst om.  En hoewel de veeleisende en sterk – maar daarom niet geheel correct – geïnformeerde klant van vandaag meer zelf bepaalt dan vroeger, blijf ik ervan overtuigd dat je als bedrijf nieuwe behoeftes en oplossingen succesvol kan creëren.  Kijk maar naar Apple’s iPod en iTunes revolutie.  Maar dat je daartoe alles moet weten over je klant en z’n behoeftes of frustraties, staat als een paal boven water.

Wat onderscheid nu BIG DATA van de dataverzameling die we nu al hebben ?  De algemeen aanvaarde definitie heeft het over de 4 V’s (in het Engels) : Volume, Velocity, Variety, Veracity.

Vertaald naar het Nederlands wordt dat zoiets als :

1. Volume : de hoeveelheid data
2. Snelheid : de datastroom en de snelheid waarmee de data wordt geanalyseerd en doelgericht gebruikt
3. Variëteit : de verscheidenheid van de informatie en hoe ze beschikbaar is
4. Juistheid : de betrouwbaarheid van je data

 

Even inzoomen op die 4 principes.

1. Volume

We spreken echt over enorme hoeveelheden, niet enkel in totaal maar ook in volume data per unieke entiteit in je totale database.  De hoeveelheid informatie over een bedrijf, een persoon, een product, de interacties met de klant, …

Het gemiddelde Belgisch bedrijf heeft vandaag ongeveer 60 Terabytes aan data opgeslagen.  Dat zal tegen 2020 doorgroeien naar 6000 Terabytes of 5,8 Petabytes of nog 6.2 miljard Megabytes.

2. Snelheid

Net de snelheid waarmee al die data geanalyseerd en behandeld moet worden om er nuttige acties mee te ondernemen, onderscheid een grote hoeveelheid data van échte BIG DATA.  Met name de exponentiële groei van netwerkgeconnecteerde toestellen in gebouwen en in de wagen en accurate “location based information” maken dat informatie just-in-time aangereikt wordt en de juiste actie ondernomen wordt.  Een toestel zoals een internet geconnecteerde thermostaat, moet rekening houden met de weervoorspelling à la minute om echt doeltreffend het binnenklimaat optimaal te regelen.  Maar moet ook weten of en hoeveel mensen zich in de ruimte bevinden.

3. Variëteit

De grote hoeveelheid data die dagelijks geproduceerd wordt, is zeer divers van oorsprong en aard.  Het gaat om tekst, afbeeldingen, video, audio, cijfers,…

Niet al die informatie is gestructureerd beschikbaar, wat een zeer groot verschil is met de traditionele database-tools, die wel steeds om duidelijk gestructureerde data vragen.

Bovendien is het bij vele soorten data zoals afbeeldingen, video en audio moeilijk om ze automatisch te kwalificeren op inhoudelijke kernbegrippen.  En dus is het lastiger om ze logisch met elkaar te verbinden.

4. Juistheid

Door de hoeveelheid data, de snelheid waarmee je ze verwerkt en de grote variëteit, is het lastig maar des te belangrijk om te kunnen rekenen op de juistheid van je gegevens.  In sommige studies antwoorden tot bijna 30% van de respondenten geen vertrouwen te hebben in de data waarover ze beschikken om bedrijfsgebonden beslissingen te nemen.  Wanneer de volumes zo groot worden dat je bepaalde processen gaat automatiseren op basis van de verzamelde en opgeslagen data, wordt het zeker cruciaal om over betrouwbare data te beschikken.

Zoals je merkt is BIG DATA een complexe materie die veel voorbereiding en strategische keuzes vraagt waar je niet licht mee mag omgaan.

Hoe je best met BIG DATA omgaat, hou ik voor de volgende post.

 

 

9 data management tips for events

Small or large companies in the event and meeting industry have to deal with data collected on their exhibitors, partners, sponsors and attendees.  Managing this data is far too often considered a pure IT-issue, solved by technical people when they install a database management software on some in-house or externally hosted server.  And then I won’t mention all those, even bigger companies, that just keep the data in separate Excel-files per edition of their event(s).

But obviously it’s not a technical IT problem, far from.  Managing data is supposed to help you get intelligence on your customers.  You need that intelligence to make clever and business accelerating decisions to improve your product offering and maintain your competitive edge and relevance in the market.  Equally, you will need that knowledge to cater for effective, 1-on-1, relevant, converting marketing campaigns.

Even without considering the data stream that comes from opening up the gates of internet and online platforms, event and meeting organizers typically already have to deal with rather large amounts of data.  Data on their prospects and customers when it comes to exhibitors, data on their prospects and customers when it comes to attendees or visitors, data on the products and services they deliver, financial data, transaction data of ticket sales, survey information, …

All of that data is constantly changing at accelerating speed.  So how do you manage that ?  Here’s 9 tips.

1. Make data management a priority

What many event organizers don’t realize today is that accurate data on vertical markets is their prime business asset.  That must be treated with care.  Make sure this data stays in top shape.  Do so by spreading quality awareness and appropriate measures and processes across all levels of your company.  Google is not about search, Facebook is not about distributing socially appealing messages and likes, they’re about providing relevant information and advertising to customers and delivering that back to their advertisers, all based on extensive data.  You as an event or meeting organizer are no different and can even be better at it if you have all the relevant data for your markets or special interest groups.  Hell, you could even start selling your data to the Googles and Facebooks of this world.

2. Procedures for quality

Develop strong and rigorously implemented and monitored procedures for handling data.  Disallow root access to all data for everyone in the company.  Implement the processes and rules in your data management software solution.  Install a role based access rights system.

3. Hire a data specialist

He/she is responsible for sourcing, buying, importing, exporting, enriching, deduplicating and analyzing data.  Marketing helps him/her, if required, with segmentation, finding potential sources of extra or new data.

4. Segment don’t separate

Exhibitors, sponsors, partners – gold, silver, bronze, whatever – visitors, attendees : they’re all customers, just with a different business value to you.  Some or many, depending on the case, will have multiple values for different event brands in your portfolio.  Don’t store them in separate databases based on their initial contact or business value.  That value might change over time.

5. Segment further

Figure out the customer information as detailed as possible.  Categorize potential exhibitors in a structured way on industry, product groups – standardized SIC, NACE but also your event specific products groups – company size, revenue size per product group, ABC-category, competitors per product group, main industry markets, focus geographical markets,…

For visitors segmentation go for interests, buying power and buying intention.  For B2B-events add company size.

Do regular data integrity checks on these segments.  Filter out all those addresses that have slipped in over time as incomplete, missing one of the segment information elements, and take the time to complete.  This information will prove valuable to you in sales and marketing efforts.

6. Develop buyer persona

From the segment information you can derive archetypes of visitors and exhibitors.  E.g. a male maintenance technician between 30-40 years old working in the chemical industry located in the Antwerp province.

For each of these archetypes try to answer these questions :

• Priority Initiatives – What causes certain buyers to invest time in my event, and what is different about  buyers who are satisfied with the status quo and don’t attend?
• Success Factors – What operational or personal results does your buyer persona expect to achieve by attending?
• Perceived Barriers – What concerns cause your buyer to believe that your event is not their best option?
• Buyer’s Journey – Who and what impacts your buyer when evaluating their options?
• Decision Criteria – Which aspects of the competing options does your buyer perceive as most critical, and what are their expectations for each?

7. Test

Test your data all the time, newly acquired data and existing data.  Over 20% of personal data becomes inaccurate after 1 year.  And that percentage grows if you take into account product data, market data,…

Do A/B-testing on your data sets.  Test purchased or otherwise newly acquired data before importing it into your CRM-database.  Use the test results to score data providers or to develop a data quality game inside your company between product range departments.

8. Build a solid data warehouse solution

Externally or internally hosted, doesn’t matter, that’s up to the manpower and skills you have available, but always look for a solid, scalable, performing, secure data warehouse solution. Set it up including a test, staging and live environment.

9. Prepare for BIG DATA

While all of the above is SMALL DATA and already a big challenge to maintain well, you must play in the back of your head with the idea of going for BIG DATA.  It will, as in many other industries, and not long from now, be of the utmost importance to event organizers.  You will want to know everything there’s to know about your exhibitors and visitors.  This information will not be available to you in clear, structured packages, like SMALL DATA.  It will have to be processed at ultra-high speeds and the volume will be huge.  This is a true challenge but when executed correctly this data will give you the ultimate knowledge of your target groups and insights needed to keep your event top of mind.

BIG DATA may seem like an overwhelming and over-anticipated monster, but with the appropriate strategy and tools, professional guidance and continuous evaluation and adjustments, you will master this too.  Big data in events : a great topic for my next post(s).