With its incredible reach, marketers have gravitated to Facebook. BRADLEY ELLIOTT, director of Platinum Seed, digs into the untapped potential of the social network’s “big data.”
Since the first friend request was made on the Harvard campus in 2004, Facebook has become the biggest social media platform in the world by far, with 1.94 billion active users, which is more than Twitter, Whatsapp and Instagram combined.
Five new profiles are created every second, and half a million comments are posted on the network every minute. It’s a huge ecosystem of data, with a wide range of content – from funny, inspiring, cute and heart-warming content, to the more tragic and disturbing. Unfortunately, brands struggle to harness the potential of Facebook “big data” – the massive reserve of data generated from this environment – to its maximum potential.
Businesses simply don’t realise how much data they collect every day and over the years. While most organisations use some of this data to make basic business decisions, like dealing with complaints, few use it to create and nurture relationships with existing and potential customers. Instead, marketers have become masters of their own demise, sending an endless stream of spam and then trying to figure out ways to increase open rates for this often unsolicited, irrelevant communication. However, marketers could very easily use available demographic data to segment their data sets to serve communication in a more targeted manner. Directing male products to males and female products to females are two prominent examples, but most businesses neglect even this basic function.
While some brands can create segments, they rarely use them at a significant scale due to the complexities involved. Being able to segment consumers based on both demographic and behavioural data is extremely powerful, but marketers are intimidated. Tapping into real-time behavioural data allows brands to keep their fingers on the pulse of where consumers are in their lifetime journey, how their preferences and habits are changing, and which products would be most relevant to them. By driving personalised communications, offers and rewards to consumers on a one-to-one level, brands can build advocacy, as well as increase purchase frequency and customer lifetime value.
Facebook has increasingly become a pay to play space, as organic reach for brand pages is now set at only 2%. Brands buy reach and engagement by promoting the content they serve to their audiences. While this can be highly targeted, reach and engagement tend to be a function of budget. The Facebook environment offers a powerful way to reach and engage customers in a much more meaningful way, but this needs to be done strategically, using the correct approach and tools. Brands that produce truly emotive content achieve higher levels of what is referred to as “share of emotion”, or content that drives people to willingly want to share content, resulting in organic reach and engagement.
Too many brands use social media to push their products on consumers, when they should be adding value through rich story-telling, helpful advice, or unbelievable facts. Most users are exposed to up to 1500 stories a day on Facebook, but an average user only gets to see about 100, which is why Facebook tries to make a user’s newsfeed as “personalised” as possible. The algorithm uses several key metrics, including the “relevancy score”, which uses hundreds of variables to control the news feed to predict what content users are most likely to engage with, based on past engagement, and the same applies to Facebook ads.
In July 2015, Facebook introduced the “see first” feature, which lets users hand-pick the accounts, whether friends or followed pages, they prefer to see first at the top of the news feed. If a user spends more time on a particular post, Facebook is more likely to show that post on friends’ news feeds and this need not be engagement in the traditional sense. For instance, people who are interested in a video might not necessarily have liked, commented on or shared it with their friends. Facebook has started monitoring other forms of video engagement, like turning on the audio, switching to full-screen mode, enabling high definition or saving a post for later viewing.
Facebooks’ algorithm is the most complex out of any social media platform and the company claims to continually change it to give users the best experience possible. Brands that produce engaging content, which is relevant to users, will have their content prioritised, but the key is to understand what content is working for the brand. In addition to great content, brands can also use brand influencers in their Facebook communities. These may not necessarily be users with the most followers, but those who drive the velocity of conversation around brand content.
Influencer software, Contunion, allows brands to identify and engage with these users and determine the overall influence of their social media communities. Continuon provides brands with key insight into the content that is getting the most engagement both on a community and individual level, allowing the brand to tailor its content plans for the biggest impact. The Continuon system considers historic and fresh data as equally important, blending it to yield more targeted insights.
With the increasing capability of machine learning and artificial intelligence, combined with large data sets, this is becoming more and more precise. Computers can now analyse millions of variables in real-time, far beyond what humans are capable of, to determine probable outcomes. Not only does prescriptive analytics predict future activity, it also recommends the best course of action for any given situation.
Which IoT horse should you back?
The emerging IoT is evolving at a rapid pace with more companies entering the market. The development of new product and communication systems is likely to continue to grow over the next few years, after which we could begin to see a few dominant players emerge, says DARREN OXLEE, CTOf of Utility Systems.
But in the interim, many companies face a dilemma because, in such a new industry, there are so many unknowns about its trajectory. With the variety of options available (particularly regarding the medium of communication), there’s the a question of which horse to back.
Many players also haven’t fully come to grips with the commercial models in IoT (specifically, how much it costs to run these systems).
Which communication protocol should you consider for your IoT application? Depends on what you’re looking for. Here’s a summary of the main low-power, wide area network (LPWAN) communications options that are currently available, along with their applicability:
SigFox has what is arguably the most traction in the LPWAN space, thanks to its successful marketing campaigns in Europe. It also has strong support from vendors including Texas Instruments, Silicon Labs, and Axom.
It’s a relatively simple technology, ultra-narrowband (100 Hz), and sends very small data (12 bytes) very slowly (300 bps). So it’s perfect for applications where systems need to send small, infrequent bursts of data. Its lack of downlink capabilities, however, could make it unsuitable for applications that require two-way communication.
LoRaWAN is a standard governed by the LoRa Alliance. It’s not open because the underlying chipset is only available through Semtech – though this should change in future.
Its functionality is like SigFox: it’s primarily intended for uplink-only applications with multiple nodes, although downlink messages are possible. But unlike SigFox, LoRa uses multiple frequency channels and data rates with coded messages. These are less likely to interfere with one another, increasing the concentrator capacity.
Ingenu Technology Solutions has developed a proprietary technology called Random Phase Multiple Access (RPMA) in the 2.4 GHz band. Due to its architecture, it’s said to have a superior uplink and downlink capacity compared to other models.
It also claims to have better doppler, scheduling, and interference characteristics, as well as a better link budget of 177 dB compared to LoRa’s 157 dB and SigFox’s 149 dB. Plus, it operates in the 2.4 GHz spectrum, which is globally available for Wi-Fi and Bluetooth, so there are no regional architecture changes needed – unlike SigFox and LoRa.
LTE-M (LTE Cat-M1) is a cellular technology that has gained traction in the United States and is specifically designed for IoT or machine‑to‑machine (M2M) communications.
It’s a low‑power wide‑area (LPWA) interface that connects IoT and M2M devices with medium data rate requirements (375 kb/s upload and download speeds in half duplex mode). It also enables longer battery lifecycles and greater in‑building range compared to standard cellular technologies like 2G, 3G, or LTE Cat 1.
Key features include:
· Voice functionality via VoLTE
· Full mobility and in‑vehicle hand‑over
· Low power consumption
· Extended in‑building range
Narrowband IoT (NB‑IoT or LTE Cat NB1) is part of the same 3GPP Release 13 standard3 that defined LTE Cat M1 – both are licensed as LPWAN technologies that work virtually anywhere. NB-IoT connects devices simply and efficiently on already established mobile networks and handles small amounts of infrequent two‑way data securely and reliably.
NB‑IoT is well suited for applications like gas and water meters through regular and small data transmissions, as network coverage is a key issue in smart metering rollouts. Meters also tend to be in difficult locations like cellars, deep underground, or in remote areas. NB‑IoT has excellent coverage and penetration to address this.
The LPWAN technology stack is fluid, so I foresee it evolving more over the coming years. During this time, I suspect that we’ll see:
1. Different markets adopting different technologies based on factors like dominant technology players and local regulations
2. The technologies diverging for a period and then converging with a few key players, which I think will be SigFox, LoRa, and the two LTE-based technologies
3. A significant technological shift in 3-5 years, which will disrupt this space again
So, which horse should you back?
I don’t believe it’s prudent to pick a single technology now; lock-in could cause serious restrictions in the long-term. A modular, agile approach to implementing the correct communications mechanism for your requirements carries less risk.
The commercial model is also hugely important. The cellular and telecommunications companies will understandably want to maximise their returns and you’ll want to position yourself to share an equitable part of the revenue.
So: do your homework. And good luck!
Ms Office hack attacks up 4X
Exploits, software that takes advantage of a bug or vulnerability, for Microsoft Office in-the-wild hit the list of cyber headaches in Q1 2018. Overall, the number of users attacked with malicious Office documents rose more than four times compared with Q1 2017. In just three months, its share of exploits used in attacks grew to almost 50% – this is double the average share of exploits for Microsoft Office across 2017. These are the main findings from Kaspersky Lab’s Q1 IT threat evolution report.
Attacks based on exploits are considered to be very powerful, as they do not require any additional interactions with the user and can deliver their dangerous code discreetly. They are therefore widely used; both by cybercriminals looking for profit and by more sophisticated nation-backed state actors for their malicious purposes.
The first quarter of 2018 experienced a massive inflow of these exploits, targeting popular Microsoft Office software. According to Kaspersky Lab experts, this is likely to be the peak of a longer trend, as at least ten in-the-wild exploits for Microsoft Office software were identified in 2017-2018 – compared to two zero-day exploits for Adobe Flash player used in-the-wild during the same time period.
The share of the latter in the distribution of exploits used in attacks is decreasing as expected (accounting for slightly less than 3% in the first quarter) – Adobe and Microsoft have put a lot of effort into making it difficult to exploit Flash Player.
After cybercriminals find out about a vulnerability, they prepare a ready-to-go exploit. They then frequently use spear-phishing as the infection vector, compromising users and companies through emails with malicious attachments. Worse still, such spear-phishing attack vectors are usually discreet and very actively used in sophisticated targeted attacks – there were many examples of this in the last six months alone.
For instance, in late 2017, Kaspersky Lab’s advanced exploit prevention systems identified a new Adobe Flash zero-day exploit used in-the-wild against our customers. The exploit was delivered through a Microsoft Office document and the final payload was the latest version of FinSpy malware. Analysis of the payload enabled researchers to confidently link this attack to a sophisticated actor known as ‘BlackOasis’. The same month, Kaspersky Lab’s experts published a detailed analysis of СVE-2017-11826, a critical zero-day vulnerability used to launch targeted attacks in all versions of Microsoft Office. The exploit for this vulnerability is an RTF document containing a DOCX document that exploits СVE-2017-11826 in the Office Open XML parser. Finally, just a couple of days ago, information on Internet Explorer zero day CVE-2018-8174 was published. This vulnerability was also used in targeted attacks.
“The threat landscape in the first quarter again shows us that a lack of attention to patch management is one of the most significant cyber-dangers. While vendors usually issue patches for the vulnerabilities, users often can’t update their products in time, which results in waves of discreet and highly effective attacks once the vulnerabilities have been exposed to the broad cybercriminal community,” notes Alexander Liskin, security expert at Kaspersky Lab.
Other online threat statistics from the Q1, 2018 report include:
- Kaspersky Lab solutions detected and repelled 796,806,112 malicious attacks from online resources located in 194 countries around the world.
- 282,807,433 unique URLs were recognised as malicious by web antivirus components.
- Attempted infections by malware that aims to steal money via online access to bank accounts were registered on 204,448 user computers.
- Kaspersky Lab’s file antivirus detected a total of 187,597,494 unique malicious and potentially unwanted objects.
- Kaspersky Lab mobile security products also detected:
- 1,322,578 malicious installation packages.
- 18,912 mobile banking Trojans (installation packages).
To reduce the risk of infection, users are advised to:
- Keep the software installed on your PC up to date, and enable the auto-update feature if it is available.
- Wherever possible, choose a software vendor that demonstrates a responsible approach to a vulnerability problem. Check if the software vendor has its own bug bounty program.
· Regularly run a system scan to check for possible infections and make sure you keep all software up to date.
- Businesses should use a security solution that provides vulnerability, patch management and exploit prevention components, such as Kaspersky Endpoint Security for Business. The patch management feature automatically eliminates vulnerabilities and proactively patches them. The exploit prevention component monitors suspicious actions of applications and blocks malicious files executions.