Modernization – Better, faster and cheaper!

Business oriented IT Managers must bring modernization to the table now because it will increase the value of IT and the value of using SAS.

Management will probably approve projects that are better (quality, aligned with strategy), faster (time) and cheaper (cost). Reaching all three parameters is difficult because higher quality usually implies increased costs and maybe delays. Faster implementation could imply worse quality or increased costs. Costs savings could imply delays and worse quality. Is it possible to have it all?

Could we have SAS running better, faster and cheaper?
Better means more robust, increased reliability, and solutions allowing for increased insight. Faster means that data can be accessible more quickly and that analysis takes less time. Cheaper relates to total cost of operation with lower hardware-, software- and maintenance costs as three examples.

Hadoop and SAS Grid
SAS customers have used the yearly user conference, SAS Global Forum, to share results after modernizing their SAS® platform, and they have proved that modernization projects can be both better, faster and cheaper.

Download White Paper

Download White Paper: SAS® Grid Computing – What They Didn’t Tell You

The main reasons for this is that modernization usually results in the introduction of clusters (Hadoop and SAS® Grid). Clusters allows for more data and increased insight, throughput and faster analysis. Hadoop and Grid runs on commodity hardware, which is more economical than a variety of servers. Fewer servers implies reduced costs.

SAS customers with old software and a variety of servers should put modernization on their agenda. I will argue that the business case is good, and that these kinds of projects will increase the value of deploying SAS within three distinct areas; It will be better, it will be faster and it will be cheaper.

Explore SAS® In-Memory Analytics, SAS® Solutions for Hadoop and SAS® Grid

Download White Paper: SAS® Grid Computing – What They Didn’t Tell You

Post a Comment

Impressions from SAS Global Forum 2016 – bigger, more connected and a strong Nordic turnout

Petter_DybvigBy Petter Dybvig, Director SAS Technical Support, Nordic Region

Having just returned from a whirlwind week in sunny Las Vegas, Nevada, I want to share with you a few of the many news and impressions from our biggest event of the year.

First of all, we were excited to welcome customers, partners and SAS employees to the largest SAS Global Forum ever! More than 5000 attendees assembled at the impressive Venetian Hotel: It is one of the largest hotels in the world and we had a fantastic venue for the forum’s 350 hours of content, 550 sessions, and many opportunities to network.

The Nordics were strongly represented with the third largest group of 250 participants, falling behind only the American and Canadian delegations in numbers.

The largest SAS Global Forum ever with more than 5000 attendees, 250 from the Nordics.

The largest SAS Global Forum ever with more than 5000 attendees, 250 from the Nordics.

Participants from all over the world had the opportunity to hear cutting-edge cases from Nordic companies such as Danica Pension, Statistics Sweden, Telenor, The Norwegian Seafood Council and Veikkaus. Our Nordic delegation could participate in a sightseeing trip to the Hoover Dam and Lake Mead, in addition to participate in a Nordic dinner. It was great to see new connections made and old ones confirmed once again!

New, more customer-oriented and connected to the IoT
So, what were the big news, we were able to experience during this year’s SAS Global Forum? For me, there were three major new announcements:

  1. We are going live with a completely new architecture, SAS® Viya
  2. The new SAS® Customer Intelligence 360 solution
  3. SAS® Analytics for IoT

SAS® Viya is a big deal because it is the new SAS architecture built for the future for both cloud and on-premise deployments. SAS is building on our unique

foundation of expertise to deliver the most powerful and adaptable analytics platform available anywhere. My

SAS® Viya – next generation of analytics

SAS® Viya – next generation of analytics

impression is that there was really good reception from customers, and I see the move to SAS Viya as a logical step due to the rapid changes in technology we are experiencing. With SAS Viya, SAS is also opening up to more third party technologies, which is great news for customers who want to leverage existing skill-sets when using SAS analytics, including languages like Python, Lua and Java. Learn more about SAS Viya


SAS® Customer Intelligence 360: The new SAS® Customer Intelligence solution 6.5, launched along with SAS Customer Intelligence 360, is a completely new product. It builds on the Customer Decision Hub concept and allows marketers to guide the customer journey as it happens. It helps marketers create a complete profile from disparage data, serving digital departments and helps optimize customer messaging and offers. The CI solution will be a great leap in helping companies increase loyalty in a cloud-based solution. Learn more

SAS® Analytics for IoT: SAS is really bringing analytics offerings to the market: IoT creates a lot of data, and long before IoT became trendy, SAS was analyzing data from sensors and other devices. With the launch of SAS Analytics for IoT, we can collect data and analyze millions of events per second. An example is to look at the idea of Smart Cities – modern cities generate all kinds of services and related data within areas like public safety, healthcare and water. This allows you to organize the data and pick out the relevant information and understand which actions to take in order to improve a specific area.

Another wealth of exciting IoT possibilities come from the Intel collaboration. SAS brings the IoT analytics offerings to market as a strategic partner in Intel’s Internet of Things Solution Alliance. Intel’s slogan is “Intel inside, smart manufacturing outside”. For example, it will be possible to place an Intel card in cars and predict issues in the fleet before failures occur using SAS software on the device in the car and provide new value added services. When you need to analyze a situation, it will be analyzed right on the chip with SAS® Software. This is analytics close to the origin of the data; it is on the edge. Get inspired

The SAS® 9.4 platform you know – with more “bells and whistles”
However – most of our clients attending SAS Global Forum were there to hear about the existing products and services they have come to know and trust, within the SAS 9.4 software suite. Both within Analytics, BI, Data Management, Risk, and Customer Intelligence there were lots of new things to experience, and the majority of the 550 sessions were centered on these. Among the highlights from the existing products were:

Analytics: SAS continues to provide analytic innovation on the SAS 9.4 platform – improving scale and speed along the way. SAS® Studio, our web-based interface that comes with SAS® Base, continues to address the analytical needs of everyone. We are releasing SAS® Factory Miner 14.1 – this product provides modern machine learning algorithms and large-scale automation models, enabling clients to distribute models to a larger user group in the company. SAS® Enterprise Miner 14.1 brings new machine learning methods and open source integration (PMML 4.2, R, Python) with continued technology advancement (for Hadoop and MapR). SAS® Forecast Server 14.1 provides a more modern web-based interface with custom code, tracking, multi-stage modelling and segmentation via demand classification.

BI: SAS® Visual Analytics 7.3 includes the HTML5 report viewer, also known as the modern viewer. It provides an intuitive, self-service approach for data preparation, analysis and collaborative sharing of results. Visually explore and interact with both structured and text data – of any size from any source – to spot patterns and identify relationships.

Data Management: I think that the main new thing for SAS® Data management is the SAS® Event Stream Processing 3.2 as well as the SAS® Data Loader for Hadoop. It is important because it opens up for SAS accessing more data, including IoT data. SAS knows that clean and correct data is key to deriving good analytical insights. With our updates focused across all Data Management technologies, we continue to focus on the entire analytics lifecycle.

Risk Management: We have streamlined products and improved scalability across the product suite.

Product demo’s and cutting edge student projects
Between the many announcements and exciting plenary speakers at SAS Global Forum, there was time to visit the many workshops and also “The Quad”, the demo area where we can go ask developers and product managers, speak with real experts about the products and the future directions, and see the future product releases. The atmosphere was great and I met many attendees who came back excited and said “oh, I learned a lot about this specific code, how to do this installation”, and so on. It is a great feeling when you can see the shared knowledge of customers, partners and SAS employees growing in real time!

I also want to mention the academic forums, where students from around the world competed with their SAS solutions based on our free-of-cost SAS® University Edition software. The winning team was from the US. This year, we did not see Nordic participation and I really hope this will happen in the future.

Once again, I have come back full of inspiration and excitement about our great SAS community and if you were there, I hope you can recognize the sentiment. If you were not able to attend, I hope you will go here to find whitepapers, presentations and much more, and we will also soon be sharing these in the local Nordic SAS Forums and SAS user groups.

Whether or not we met in Vegas, there is another opportunity in the fall when the next line of SAS Nordic Forum events roll around. They are set for September 29 in Norway, October 4 in Sweden, October 5 and 6 in Finland, and October 13 in Denmark.

And for those of you that already want to start planning for the next SAS Global Forum, it will take place in Orlando, Florida from April 2 to April 5 2017.

Hope to see you there!

Post a Comment

Copy cats, gamblers or entrepreneurs?

Did you find that special card of diamonds in the card deck if you were gambling in Las Vegas? Did you find that golden egg? Did you manage to keep up with everything that was happening almost 24/7?

The program at SAS Global Forum was set from morning to late hours, 3 days in a row; Networking opportunities, sessions and keynotes, demos and Talking Points gave us the chance to hear other people´s experience in using SAS, and the ability to discuss with SAS Experts and be part of very exciting new SAS launches.

We also spent time to network and establish new relations, with Nordic and global colleagues, customers and partners, spending valuable hours socializing and sharing knowledge and inspiration, in and around the amazing Las Vegas.

I speak for myself when it comes to finding those golden nuggets. They were there. I learned a lot during the conference and got to know many new, great customers, which is very valuable for me personally.

Busy schedule at SAS Global Forum

Busy schedule at SAS Global Forum

Your experience - Inspiration from your colleagues
What did you experience at SAS Global Forum? What did the presentations and findings, and new ways of doing things mean to you? Moreover, to your manager and organization, that has invested in your presence at the conference? I believe one positive effect is that you will work differently when returning after SAS Global Forum. Do you agree?

Driving your business forward means challenging the normality’s and to adjust the governed way to do things. Understanding the way other people think, does that mean you are copying their work? Alternatively, does it prevent you from repeating their flaws and potentially doing things faster and more accurate using your new knowledge? I believe you will find this out yourself the next days, weeks and months, and that it will be much easier to convince your manager next year letting you go to Orlando, Florida and experience SAS Global Forum 2017, with the new insights and great work coming from you.

I guess you also learned to know new people that you will stay in contact with, both from SAS Institute, and other organizations. Without the opportunity to go to SAS Global Forum you would not had the same chance to network. Valuable to you personally and professionally, but also of high value to your organization in order to solve challenges in your daily work.

Presentations and recordings
One thing that troubled me during the conference, and probably some of you attending, was that many of the presentations I wanted to attend occurred in parallel. Well, I have great news for you. You can now download all proceedings and papers available from the conference, via the application on your phone or the SAS Global Forum web site. You can also see presentations and interviews on video. You will not miss out after all!

Nordic presenters at SAS Global Forum
SAS owes a huge thanks to everyone who contributed with presentations at this year’s conference. Without them sharing their knowledge there will not be a conference next year. Now we are building the program for SAS Forum and SAS Business Forum for all Nordic countries, taking place this fall. If you want to share your SAS knowledge and inspire others, submit your brief abstract (max 300 words) of your idea to Program Committee lead Georg Morsing: before May 25.

Hoping to see you all this fall at our Nordic SAS Forums and next spring at SAS Global Forum 2017! Meanwhile, you could read more on our new, exciting launches:

SAS® Viya

SAS® Customer Intelligence 360

SAS® Analytics for IoT


Post a Comment

What happens in Vegas should not stay in Vegas?

SAS Institute has its annual user conference in Las Vegas, SAS® Global Forum, and it is the largest conference ever with more than 5 000 attendees. Users and analysts from all over the world hear about use cases from all industries - from customers, partners and SAS. The demo area displays new technology and solutions, and our important consulting partners are aligning their offerings with SAS. We also experience great opportunities to network with our peers worldwide.

What happens in Vegas is all about analytics, and how analytics in action creates value in business processes cross industries. Industries like banks and insurance companies have used analytics for years, but analytics is rapidly moving into other industries. It is happening now, it is happening fast. During the conference first day, we heard about real use cases that creates tangible business value. Not tomorrow, but today!

More emphasis on capabilities
 Lessons learned from my first day in Vegas is probably that we should emphasis less on technology but more on capabilities, and that we should use the right tool for the job. SAS Institute has a strong offering because SAS provided capabilities to succeed with analytics. However, an important new capability from

Presenting SAS® Viya – next generation of analytics

Presenting SAS® Viya – next generation of analytics

SAS is open API’s and support for Hadoop distributors. This enables innovation because developers can work outside SAS with other technologies and access capabilities provided by SAS, like third party applications calling SAS analytical models.


SAS in the market for 40 years and more to come
SAS celebrates its 40th year anniversary this year, and the company has grown profitability every year. Some analysts could argue that the era of SAS is over and that new entrants will take lead and influence the next 40 years. I am willing to take a real “Las Vegas bet” and bet my salary that they are wrong.

My argumentation for this is that the new capabilities in SAS will enable rapid innovation with better support for the analytical life cycle. SAS will continue to support the user community with necessary analytical capabilities and respond to requirements. SAS business- and technology partners will continue to deliver high value to our clients.

Presenting SAS® Viya – next generation of analytics
Today SAS announced a new modern architecture that is future ready, SAS® Viya. SAS solutions on Viya attract and retain the next generation of analysts and

enable everything from start-ups to large enterprises with actionable analytics.

Which competitor can catch up and compete with SAS? SAS is, without discussion, the market leader, and the company will continue to lead way for the next 40 years. Businesses that rely in SAS will get a competitive advantage over companies that struggles with their analytical strategy.

Interested in learning more about SAS® Viya?
SAS® Viya – a new, open architecture – built for analytics innovation

Press release on SAS® Viya

Post a Comment

How Sports should engage with Analytics

If you don’t believe Analytics and Sports go together you may want to stop reading here. Or you may want to start with validating my previous reasoning on "Why Sports should embrace analytics".  If you find you are past the “Why?” you may instead think that the title above doesn’t call for a page long article, that Nike’s slogan “Just do it” would be sufficient as a recommendation. I can see that point of view, I used to think that surfacing information that was previously unknown would be enough to trigger engagement, buy-in and action. However, more than once I’ve found that exposing a sports management team to information they never saw before based on data they didn’t realize existed leads to nothing following the meeting. Even though they clearly both see and express how these things would be valuable to them it doesn’t necessarily move forward from there. So, assuming that the question “Why?” is not an issue, let’s take a look at “How?”, how do you engage with analytics in order to make it operational and contribute to the team?

1. Be relevant.

Realizing this may at first sight qualify as a true “no-brainer”, it still can’t be pointed out often enough. Relevance is everything and it applies to information, communication as well as timing.

Let’s start with information relevance, what information are you presenting? In sports especially, there are so many things that you can do with statistics and analytics that are interesting, and maybe even entertaining, and they can lead to amusing discussions, but if you want people to truly engage it is essential that the information is relevant, something that clearly aligns with an agreed objective. For the sake of illustration, let’s use hockey face-offs as a very basic example.

Possessing the puck is important, therefore face-offs are important, hence measuring and knowing different players’ face-off percentages are valuable. 50% is good, higher is of course better, 60% is excellent, above that is rare and exceptional. But players’ faceoff percentages are not consistent around the rink, they typically vary considerably between different circles on the ice. So the detailed information below adds relevance since it more clearly indicates who to have or not have on the ice pending the situation.

Crucial face-off in your defensive left circle, you may want somebody else in there, right? As the British Rowing team phrased it, if it provides an answer to the question “Will it make the boat go faster?” the information is relevant.

Source: Twitter account @JonBlomqvist


Let’s move to Communication relevance. How to visualize your findings. The human brain remembers pictures. Look again at the image above, it takes a few seconds only for everyone in the room to grasp the context and it’s easy to recall from memory. And as you talk about the findings and the implication, make sure you use a language that is understood and keeps the listener on track, words like “odds” and “probability” will probably work while “standard deviation” and “r square value” will likely not.

Timing relevance relates to understanding when the information adds value. In connection with practice or with game? When evaluating, preparing or executing? Understand the process and add value to it by timing the information. If it’s not relevant to the situation it will become noise, no matter how correct and valuable it may seem to you. And insights not put into practice remain only insights, nothing more.

As you move forward from known areas and statistics to new and more advanced analysis, as long as you stick to relevance in information, communication and timing you can’t go wrong.

2. Form a team.

Engaging in Analytics is not something for one person to do in isolation. To be successful the whole staff needs to embrace analytics, from GM to assistant coaches, they have to be involved, one way or another. Even if one person has all the skills required, which is rare, you above all need collaboration and communication, validation of findings and different perspectives in order to achieve a joint trust of results and execution thereupon

A sports team that follows these guidelines on “How” will gain insights that will lead to actions that will make a difference.

And by the way, if you’re not involved in sports but work in business or public service, this same answer to “How?” when embracing Analytics applies in order to gain success. Be relevant with regards to Information, Communication and Timing and Form a Team.

At the Data Innovation Summit event I laid out some of these thoughts in my main stage presentation and in a wider context for a few minutes in this  video. Your feedback and thoughts are highly appreciated.

Previous blogs:
Why sports should embrace analytics


Post a Comment

Merverdiavgiftkarusellen – Missing Trader

«Merverdiavgiftssystemet er basert på tillit. Det er de næringsdrivende selv som deklarerer avgift som skal betales eller utbetales, og dette gir betydelig risiko for både ubevisste og bevisste feil hos den avgiftspliktige». Denne forumuleringen er sakset fra rapporten om Riksrevisjonens undersøkelse av skatte- og avgiftsmyndighetenes kontroll av merverdiavgift, og jeg ønsker i dette innlegget å ta for meg svindel med merverdiavgift, med formuleringen som utgangspunkt.

Det store sorte momsmarkedet
Flere undersøkelser og aksjoner fra skattemyndighetene viser at svindel med merverdiavgift foregår i betydelig omfang innen EU. Når vi ser nærmere etter, ser vi at utformingen av EUs merverdiavgiftssystem åpner opp for flere former for svindel innen avgiftssystemene.

Vi kaller disse for merverdiavgiftkaruseller. Disse er arnested for uriktig momsfritak på grenseoverskridende transaksjoner med muligheter for å utsette betaling av moms ved import. Dette rammer i hovedsak stater med forholdsvis like momsregler, m.a.o. medlemstatene i EU hvor disse type avgifter er harmonisert.

Momsgapet innen EU er beregnet til svimlende av 200 milliarder Euro, dvs. i overkant av et norsk statsbudsjett. Det antas at om lag 20% av dette beløpet hvert år forsvinner inn i kriminelle nettverk.

Utformingen av EUs merverdiavgiftssystem åpner opp for flere former for svindel innen avgiftssystemene.

Utformingen av EUs merverdiavgiftssystem åpner opp for flere former for svindel innen avgiftssystemene.

Hva er Missing Trader?
Missing trader, eller momskarusellen, rapporterer at det er eksportert varer mellom bedrifter i forskjellige land. Her gjelder det imidlertid en fiktiv rapportering hvor deltagerne lar varen gå «i sirkel» mellom avtalte bedrifter, derav navnet karusell. Hvorpå varene, utbetalt moms og aktører plutselig forsvinner.

Et eksempel kan være at en merverdipliktig bedrift i Danmark kjøper datamaskiner i Tyskland, for så å rapportere varen eksportert uten at moms blir betalt. På papiret selges varen først videre til Frankrike og deretter til Luxembourg for til slutt å ende opp i Belgia. Deretter forsvinner selskapet som kjøpte varen, eller varen forsvinner ut av systemet, uten at momsen faktisk har blitt betalt. Selskapene som har deltatt i karusellen har avtalt dette spillet, og sitter igjen med momsrefusjonen. Dette beløpet sendes kanskje til forretningsforbindelser utenfor EU. En slik avgiftsunndragelse er med andre ord svært krevende å ordne opp i når varen og aktørene er borte.

Hva med lille Norge?
Norge er knyttet til EU gjennom EØS, men erfaringene fra EU er relevante for Norge selv om vi ikke har et tilsvarende momsfritak for eksport. Eksempel på dette, gjenfortalt fra Økokrims trusselvurdering, er historien om Tollvesenet som stoppet 60 containere med totalt ca. 1 300 tonn hvitløk som skulle importeres til Europa via Norge. Vi hører her om at import av hvitløk til Norge er avgiftsfritt, mens tollsatsene inn i EU er 1 200 euro pr tonn.

Alt handler om hastighet
Som du forstår er denne type svindel i sin natur kompleks, noe som gjør det krevende for myndighetene å mobilisere raskt nok. Når tiltak gjennomføres, er varene, aktørene og pengene forsvunnet. Et angrep mot denne type svindel krever derfor et høyhastighetssystem for å kunne identifisere potensielle sammenhenger mellom kjedene av virksomheter som er involvert, mens spillet ennå pågår. En deteksjon av falske nettverk krever lynrask, avansert analyse med stor presisjonsgrad på store og ofte ustrukturerte data. Det er her big data kommer inn.

SAS har hjulpet myndighetene i flere land
Momskarusellen er høyhastighetssvindel hvor evnen til å kunne stoppe momsunndragelsen er proporsjonal med deteksjons- og responshastigheten. Løsningen SAS Institute leverer for å hjelpe myndighetene med å oppdage selskaper som, med høy sannsynlighet er involvert i organisert merverdiavgiftsbedrageri, benytter en hybrid analysemodell. Løsningen heter SAS® Fraud Framework for Government. En hybrid modell kombinerer scoringer gjennom flere analysedimensjoner og skaper på denne måten et holistisk overblikk over situasjonen. Dette gir myndighetene muligheter for snarlig oppdagelse, fra den første mistenkelige omsetningsoppgave eller annen mistenkelig aktivitet.

Vår hybride modell kombinerer nettverksanalyse, forretningsregler, prediktiv modellering, tekstanalyse, beslutningstrær og anomali gjenkjenning. Kombinert med riktig innsikt på brukersiden vil et slikt angrepsvåpen kunne identifisere og monitorere nettverk hvor det med overveiende sannsynlighet foregår aktiviteter som vil ende opp med avgiftsunndragelse. Modellene er svært nøyaktig.

Tilbakemelding fra prosjekter
Resultater fra Belgia og Nederland viser omfattende reduksjon av momskarusell svindel. Avansert analyse har vært nøkkelen i å oppnå dette resultatet. Modellene er ikke bare svært nøyaktige (80% sanne positiver), men de gir også forvarsel fra de første merverdiavgifterklæringene. Erfaringene fra Belgia og Nederland tilsier derfor at EUs visjon om at organisert momsbedrageri skal tøyles burde være oppnåelig. Vår løsning er også tilrettelagt for internasjonalt samarbeid ved å identifisere mistenkelige selskaper på tvers av land. Som kjent kjenner økonomisk kriminalitet ingen landegrenser.

Lær mer om SAS® Fraud Framework for Government


Post a Comment

Konkurransekriminalitet – Teori og praksis

For mange av oss er det først når vi hører om millionbeløpene innen konkurransekriminalitet at vi slår opp øynene. Ser vi nærmere etter, finner vi også pris- og konkuransesamarbeid, og annen kartellvirksomhet, på mindre oppdrag og avtaler som over en lang periode genererer store beløp.

Kartellmedlemmer samarbeider om pris og styrer det hele gjennom å fordele markedet, kunder og oppdrag mellom seg. Vi møter det i media bl.a. som oppslag om synkroniserte drivstoffpriser, unisone kassaapparat innen varehandel og virksomheter innen anlegg og samferdsel.  Fellesnevner for mye av denne type økonomisk kriminalitet er utfordringene knyttet til bevisførsel og dertil få rettskraftige dommer.

Prissamarbeid og rabatter
Innen konkurransekriminalitet er det kunden og samfunnet som er de store taperne, gjennom at konkurransemekanismene i markedet blir satt ut av spill. Vi har også sett at bransjeorganisasjoner har bidratt til ulovlig samarbeid gjennom å vedta eller promotere konkurransebegrensende retningslinjer, eksempelvis gjennom ekskluderende standardisering eller «rådgivingspakker» knyttet til prisingsstrategi og rabatter. Dette fører til at myndigheter og konkurrenter «stenges ute» (jfr. konkurranseloven §10 og EU’s Romatraktat artikkel 101- 109).


Omfanget på prissamarbeid, i enkelte bransjer, gjelder opp mot hvert femte kundeoppdrag.

For beviselig prissamarbeid anbefaler EU bøter på inntil 10% av omsetningen. I noen tilfeller aner bransjen selv at det foregår prissamarbeid og at omfanget i enkelte bransjer gjelder opp mot hvert femte kundeoppdrag. Anslag på mørketall mangler i stor grad innen konkurransekriminalitet og blir kun spekulasjoner.

4 ulike profiler
For å angripe konkurransekriminalitet analytisk må vi strukturere og modellere problemstillingen. Til dette kan vi benytte profiler. I vårt tilfelle finner vi fire arketype profiler; «pris», «marked», «anbud» og «leveranse». Disse fire parameterne gir til sammen en uærlig aktør muligheter for å estimere inntekt- og fortjenestepotensialet, gitt at aktøren kan kontrollere de fire parameterne f.eks. gjennom ulovlig konkurransesamarbeid.

Prisprofiler; Pris og konkurranse, knyttet til prisstrategi og rabatter, blir ulovlig prissamarbeid når konkurrenter har kontakt vedrørende prissetting og kommersielle forhold.

Markedsprofiler; Ulovlig markedsdeling oppstår når konkurrenter avtaler å dele marked, kunder, oppdrag eller domener mellom seg, eller avtaler produktspesialisering som tilnærmer monopol.

Anbudsprofiler; Ulovlig anbudssamarbeid innebærer at bedrifter samarbeider om entreprisekontrakter og anbud, f.eks. priser og vilkår før tilbud gis. En kompleks variant på ulovlig anbudssamarbeid er å spotte situasjoner med anbudsrotasjon, fiktive anbud eller at leverandører avstår fra å gi tilbud. International markedsadgang med komplekse selskapsstrukturer øker vanskelighetsgraden for å avsløre konkurransekriminalitet i anbudsprosesser. Hvordan er for eksempel underleverandører priset inn og med hvilken risiko?

Prosjektprofiler; Når virksomheter gir felles tilbud i tilknytning til leveranser i prosjekt, vil ulovlig prosjektarbeid oppstå når de egentlig må kunne betraktes som reelle- eller potensielle konkurrenter, i forhold til domene eller produkt. Slikt samarbeid vil også kunne rammes av konkurranseloven hvis det utveksles nøkkelinformasjon om for eksempel kalkyler og modeller. Altså både strategisk informasjon eller andre beskrivelser utarbeidet i tilknytning til oppdraget.

For å finne indikasjoner på konkurransekriminalitet kan vi score aktiviteter langs de profiler vi har beskrevet. For prisprofilene kan vi for eksempel finne mønstre i pristilbud eller rabatter blant deltagere i det vi innen spillteori kaller variabelsum-spill, dvs. en risikoscore for et mulig samarbeide til felles fordel for kartellmedlemmer over tid. Eller for eksempel innen anbudsprofilene å risikoscore anbudsrotasjon. Nash likevekt og Ikke-Pareto optimalitet vil kunne indikere anbudsrotasjon.

For å bygge slike modeller i praksis benytter vi scenarier og risikofaktorer for å for eksempel risikoklassifisere leverandørens pristilbud, rabatter, markedsandeler, geografiske domener og produktspesialisering. Hver av disse risikoscorene på aktiviteter er kanskje ikke så høye, men når vi kombinerer resultatene, for eksempel i forhold til tidslinje eller domene, vil vi kunne se mønstre som indikerer konkurransekriminalitet. Våre erfaringer med denne metoden, og tilsvarende analyser av mønstre, ved bruk av løsningen SAS® Log Analytics er gode, blant annet innn finans- og helsesektoren.

Implementering av denne typen system kan ikke bare identifisere lovbrudd, men virker også preventivt gjennom at aktørene er innforstått med at avvik kan generere alarmer og videre etterforskning.

Les mer om løsningen SAS® Log Analytics

Post a Comment

Public fraud – cheating in taxes and social benefits

Public fraud means stealing from the pockets of us all, benefiting only a few dishonest individuals. The more fraud there is, the more we need to pay to cover these losses.

Public fraud means stealing from the pockets of us all, benefiting only a few dishonest individuals. The more fraud there is, the more we need to pay to cover these losses.

Nordic Countries are often considered as the best examples of welfare states, in which the economic and social well-being of the citizens is top class. This means various social benefits in the different stages of the life of a citizen, especially in the case of hardships, such as unemployment or sickness. On the other hand, this is enabled with efficient taxation, in which we all need to put our two cents (or more likely a bit more) to facilitate the sharing of wealth to those people and situations when it’s needed.


In both sides of the equation, there is a chance of misbehavior. Taxes can be avoided or mitigated. Social benefits can be gained without a legitimate purpose. Unfortunately, there are free riders in the welfare states, who violate the common rules by either not fulfilling their liabilities or getting benefits groundlessly.

From “fine-tuning” to organized tax fraud

The most common way of tax fraud is giving the tax office false information, which decreases the amount of taxes to be paid. Typical examples are not reporting all the incomes and exaggerating deductions due to work-related expenses. These may be related to rather modest amounts of value, but the volumes can be high, because this kind of “fine-tuning” can be carried out by any individual.

In the other end of the spectrum are organized criminals, who have more efficient ways to benefit from the system. An example is value-added tax carrousel also known as missing trader fraud. This assumes a number of companies that are located across a border between countries and selling goods to each other in a long chain. The carousel takes advantage of the VAT refunds paid by the government when the goods cross the border. At some point, one of the traders in the chain claims for the refund, but does not pay the VAT, and then disappears. The carousel can then continue to operate for ages, rotating the same goods between the countries, and netting the amount of unpaid VAT every time. This has been a serious problem especially in UK, but it is carried out in other countries as well.

There are various types of tax fraud in between these two ends. Typical ones are also falsified receipts by companies.

Social benefits misused

Social benefits mostly aim at helping individuals in situations in which they require financial support due to the circumstances in order to gain acceptable standards of living. It is possible to fake such circumstances or continue to get benefits even though the circumstances change.

Regarding social benefits, the most common fraud types are related to unemployment and housing. Typically, a person claims for unemployment benefits despite working at the same time. Another common fraud is false reporting of the people living in the same household, which may lead to gaining benefits, which the total income of the household would not assume.

There is also some amount of fraud in benefits paid from sickness. Some of them are related to fake permanent disabilities and some to fake sicknesses during employment. When the person is employed, it should also bore in mind that fake sickness hurts not only the social benefit institution, but also the employer.

All in all, social benefits fraud accounts for serious financial losses to the government. As an example, the Social Insurance Institution in Finland has approximated a loss of millions of euros every year.

Analytics to the rescue

There are various ways of identifying and managing public fraud. Tax offices are actively collecting tips about tax frauds from the public. The government organizations are also actively sharing information, which helps social insurance institutions reveal the fraudulent cases, for example in which the income of a person are too high for certain type of benefit. However, this kind of fraud detection is typically carried out ad hoc, and is very time-consuming and costly. Furthermore, this way the fraud could have already been carried out for a while, until detected.

Modern information technology can also help a great deal when tackling fraud. Current data integration technologies help in automating the data collection for analysis. Analytical methods help in finding anomalies in the behavior of an individual compared to his/her peer group, for example. Predictive analytical models can be used in modelling, what are the typical features of a fraudulent activity or a fraudster, which leads to smaller target groups for investigations. Network analysis helps in linking people and organizations together, which is of an utmost importance when examining complex, organized fraud cases which involve dealings between various stakeholders. Many of the analytical techniques can create automated alerts when something statistically suspicious is carried out in the tax or benefit scene. There are also various ways to visualize the phenomenon, which decreases the time from suspicion to making the decision whether fraud has actually been carried out.

All for one and one for all

The mentality should be “all for one and one for all”. If one is getting more and giving less than he/she should, this noble idea is not fulfilled.

Luckily, as mentioned, there are more and more sophisticated ways to tackle public fraud. These methods are bit by bit adopted by government organizations that are developing their analytical capabilities. As a positive example of applying modern analytics, Special Tax Inspectorate cut down the VAT carousel fraud in Belgium nearly 98%, resulting in savings of more than a billion euros yearly. In UK, HM Revenue & Customs was able to collect approximately 7 billion euros more taxes with the help of analytics.

I will look forward to the similar success stories in the Nordics. After all, preventing public fraud is one of the key elements when maintaining our position as the welfare states.


Read more insights about Fraud and Risk.

Post a Comment

Why sports should embrace analytics

With the wide spread Moneyball success story you may think this question is already answered.  I for one used to think so. However, after I’ve enthusiastically tried to use the Oakland A’s success story as a metaphor and inspiration for applying analytics, with mixed success, I think it’s worth elaborating just a little on “Why”. What is really the reason for sports to embrace analytics?

Recently I discussed the current analytics trend in sports with one of the most successful players and GMs in Swedish Hockey. We spoke about the emergence of “sports analytics” and “fancy stats” and how “Corsi”, “PDO” etc. have started to become widely used terms in the world of hockey. I asked “With all your past experience and success, what’s really your view on this new hype?” Without hesitation he said: “What will help us become successful in the future? Practice even more? Double the off-season time in the gym? I don’t think so. The answer has to lie in understanding what really makes a difference, that’s what I expect analytics can do for us.”

To understand what makes a difference

Meeting that expectation, that’s what analytics did for the UK Olympic rowing team. Helping them understand what really makes a difference, helping them always give a fact based answer to their ever present question “Will it make the boat go faster?” When I a couple of weeks ago had the privilege of listening to Ben Hunt-Davis, one of the UK Rowing gold medalists from the Sydney Olympics, he explained what made them go from a championship participant to a Champion, from being a contender to becoming a winner.

Ben’s story was as expected about determination, focus, and their willingness to “give it all”, to take and accept the pain associated with the road to success. For years he and the team practiced ruthlessly, 49 weeks of the year, up to 21 sessions a week. For years they pushed on, convinced that if they just worked harder they would in the end succeed. What they found was that they did indeed deliver very consistent results, but finishing 6th to 8th in international competition, year after year, was not the result they were aiming for.

So Ben said they had three options. They could keep on and get the same results. Or they could simply give up. Or, they could take a step back and start looking at things in a very different way. A way to get a different result. They decided for the latter. And in the end they went from trying harder to practicing smarter. From doing more of the same to doing the right things. Using an Analytics approach helped the team to see beyond the results, it helped them measure the performance, and more importantly, to understand the correlation between performance and results. It gave them a more fact based answer to their ever present question “Will it make the boat go faster?” On the starting line of the Sydney Olympic final their boat held the youngest, least fit, least experienced rowing team, but that boat with that team was the first to cross the finish line, giving UK that gold medal again for the first time since 1912. Analytics raised their probability of winning, it changed the odds in their favor.

To change the odds in your favor


This very same approach made Detroit Red Wings’ team captain Nicklas Lidström the world’s best ice hockey defenseman, winning four Stanley Cups, the Olympics, World Championship and voted NHL’s best defenseman seven times. Anyone could see he wasn’t the strongest, not the most technical, not the hardest shooter, best tackler or swiftest skater. His road to success was to understand the odds in playing every single situation so he could minimize the opposing player’s potential, to know the odds of different choices before going out on the ice. Analytics. Not until after his career was this approach documented. And not just documented, the findings came out as a first scientific article by PhDs Carl Lindberg and Jan Lennartsson with Lidström as co-writer, explaining how the so far intangible term “game intelligence” can actually be put into the realm of science, something that can be explained. And how it can be put into practice.

But wait, there’s more

So then, the answer to “Why” is raising the probability of winning? Yes. But I would maybe claim that there’s an even more important reason why sports should embrace analytics. I believe it can be of even higher value to make sure to take fact based decisions when losing. When results are poor, emotions go high and media, sponsors, fans and board members all have (different) opinions on what needs to change. When this storm influences and even drives big decisions around firing and hiring of coaches and players. Without having embraced analytics this is where things can really go overboard. For two losing teams there’s a big difference between one that continuously evaluates and communicates performance and how it correlates with the results and one that looks solely on results.

From the top of my head I can think of at least two recent team examples where it did go overboard. But I also know of a sports team that is in the midst of a losing season and normally would have been the victim of continuous internal and external opinion storms, but they’ve stayed on course, knowing that this is a journey. One of the coaches expressed how they would never have managed to keep out infected opinions and energy drains, keep up the team spirit and the aim going forward if they had not stood on analytics. Analytics giving them facts and understandings of their performance, how it relates to the results and thereby giving them the understanding what to change and work on to improve their results over time.

So there lies the answer to the headline’s “Why”: To understand the relationship between performance and results over time in order to stand firmly on the foundation that the decision taken in a given situation was the best possible.

And by the way, if you’re not in sports but have a role in business or public service, this same answer to “Why embrace analytics?” applies. To know that the decision taken in a given situation was the best possible.

PS. I will be giving a presentation about the Why, How and What around Sports analytics and how it relates to other businesses on 22 March at Data Innovation Summit in Stockholm.

Post a Comment

Successful Industrial Internet implementation requires a broad view

How should a company begin implementing the industrial internet in their business? What are the methods by which companies can ensure that they are on the cutting edge of the on-going revolution?  The answers vary depending on the company in question, but the main thing in the increasingly demanding and competitive industry sectors, is to have an open mind and a willingness to try new things. You can find concrete examples of how it works for both the clients and the companies in the automotive industry, as well as the steel industry, says Oscar Lindqvist, Senior Advisor at SAS Institute.

In the recent years the customers in various industries have become more demanding due to increased competition and the troublesome market situation, and this brings new challenges on how companies can answer these demands.

“Companies can differentiate themselves from their competitors and establish a clear competitive edge by being the first in implementing new solutions to these new challenges”, Lindqvist says.

That could be monitoring your whole manufacturing process to ensure that everything that can be done to make the process more effective is taken into account. It could also include making the maintenance of a fleet of trucks more intuitive and predictive. There is a large array of ways you can add value to your business.

A holistic approach to quality control

In the steel industry, it is of utmost importance, that the final delivered materials meet the quality demands of the clients. If the quality doesn’t fulfill these demands, an extensive process begins.

“There’s the factor that steel doesn’t have a best before date, so some time might have passed from the date of the actual manufacturing to the moment, when the company begins to analyze the reason for the faulty batch”, Lindqvist says.

“This process of finding the root cause can take months and in the meanwhile put a serious dent in the output of the plant and affect the whole supply chain, as well as other clients, who now have to wait an additional period of time to receive their orders. This takes its toll on the value received by both the customers and the manufacturer in question.”

The Industrial Internet applications offer a solution to this by making it possible to monitor the whole manufacturing process, and gather information about it.

“You can for instance see what the temperatures were in different parts of the process, who was working on the batch that was manufactured and where the raw material came from.”

By analyzing this collected data, you start to notice connections relating to different circumstances in the process and begin to see why certain results surface, and make adjustments accordingly.

Better service through collected data

Another good example of applying the ideas of the Industrial Internet to gain a competitive edge comes from the field of automotive industry. A North European truck manufacturer has begun collecting data from routine truck maintenance in thousands of service points globally, and sending it all for processing and analyzing in their headquarters. By doing this the company can determine the remaining life cycles of certain components in the trucks and determine how long they last – or should last.

From there the processed information continues on to product development, where it can be used to make the manufactured trucks better. The information is sent back to the service points to improve the maintenance. With this data, the manufacturer can offer services and maintenance that add extra value for the clients.

“The service points can offer a campaign on brakes that relies on the expected life cycles of the part and offer the clients deals in the service points”, Lindqvist says.

The business benefits grow according to how largely they utilize these possibilities. The small streams form a large current when you’re talking about the uptime of an entire fleet of trucks. For large companies the yearly savings can be counted in millions.

Think outside the box

Both the examples mentioned earlier apply a somewhat unconventional way of approaching the businesses in question, at least from a traditional point-of-view. This might mean for example taking advantage of the large amount of information crunching start-ups.

“Start-ups are agile and have new thoughts on running businesses, so they can come up with innovative solutions fast”, Lindqvist says.

In other cases businesses can set up events to allow for a different view of thinking about the business. Recently, hackathons have become a prominent option for developing new solutions for the needs of companies. Having these kinds of events and having smaller companies participate in the development of new approaches to the business makes the whole process more agile. It also makes it possible to come up with completely new solutions to optimizing the work flow of the production plant.

“The challenge in having these kinds of approaches comes from the fact that often the added value doesn’t show immediately, instead it requires a mindset akin to scientific research from the companies willing to develop new solutions. For companies such as Google and Amazon this is easier, as their core business isn’t as clearly defined as in traditional industries. For industrial businesses it’s more important that the approaches they develop remain close to the core business.”

Originally posted on

Get more insights on Internet of Things.


Post a Comment