Immetrica


News
How Immetrica’s personall omnimeasurement can help pay-TV and OTT operators
25 August 2020
The story is the same all over the world: as analogue terrestrial transmission is replaced with digital, pay–TV penetration declines, in favour of one or more much less expensive OTT services and terrestrial television with the suddenly perfect picture quality enabled by digital transmission and its integrated error correction. This happens everywhere, albeit at different speed. The emergence of connected/smart television sets, at prices only slightly higher than unconnected ones, further contributes to this by giving standalone OTT services a direct pipe to what is often the best big screen in the house.
On top of this, pay-TV proved highly dependent on live sports, and without them during the Covid-19 pandemic it suffered several normal years’ worth of contraction. Not all who unsubscribed will return once the crisis is over and sports and incomes resume.
Cord-cutters are scary enough to pay-TV. But they are followed, as is abundantly clear now, by a generation of cord-nevers, whose members do not understand how one could tolerate watching on someone else’s schedule, or always on the same screen in the same place, or the antediluvian concept of disputes over the remote control. Television has been liberated. It will not be stuffed back into the bottle.
Of course, television distribution is not the only business of these companies. They are usually also the Internet providers, so make money on that even when they lose TV subscribers to OTT and digital terrestrial. And they sell phone service. However:
  • • The Internet is too important to leave to a natural monopoly, and fixed 4G and 5G networks are the big threat, because while slower than cable, they are fast enough. Netflix didn’t sink a large fortune into optimising network throughput for nothing. Public all-city WiFi networks and WiMax further contribute. Television adds substantially to revenue per subscriber and the double play keeps clients from leaving. It is impossible to let television go and coast on Internet delivery.
  • • Wireline telephony is increasingly irrelevant to people’s lives, while reselling another operator’s mobile phone service is a low-margin business.
What are you, as a manager of a well-run pay-TV operator, to do? First, count your advantages and make the most of them. There are a few:
  1. A. Start your own OTT/SVOD. Fair enough, but that generates some of the same problems, as discussed below.
  2. B. Prevent unsubscription (which used to be called “churn” back when leaving customers were substantially replaced with new ones; that time has passed forever). But you know that they unsubscribe only when they do, and they are extremely difficult to get back. If only you had early warning of impending unsubscriptions, you could reduce the price, upgrade the channel tier, bundle in a sports championship at no charge…
  3. C. Be stronger in retransmission and content negotiations, so know more about your customers’ use of licensed channels and programming. Will a household unsubscribe if you do not renew a specific channel? Even if you have a working RPD service, you cannot tell. You’ve heard about the major Asian operator who had to drop much of its channel lineup, including exclusive rights to a key national sport, which were immediately picked up by a competitor—the fire sale of a company’s most valuable assets being a sure sign of a flat spin into the ground. You don’t want this to happen to you.
  4. D. And the big one: you have a large subscribing household base, and the insertion minutes, one or two per hour, in which you can sell advertising do add up. But the CPM you get from them is pathetic. You know that coviewing is common at around 1.2 average viewers per viewing household (VPVH) in the US, but you cannot quote this higher number with your household-level RPD data. And you are interested in addressable advertising, because it attracts much higher CPMs, but there’s no way to target your households because you don’t even know who lives in them—or almost anything about them. Numerous firms are selling various proprietary solutions but you get the sense that they’re all based on a large number of possibly unwarranted assumptions.
personall is a smartphone-based comprehensive TV/DVR/VOD/OTT measurement system, able to recognise playout on any device, fixed or mobile. How can it help?
D. A personall panel of a relatively small number of subscribers is sufficient to produce both VPVH and demographics. Your advertising inventory is suddently demographically targetable. This is the simplest solution, which delivers results without any connection to your RPD except through the channel lineup and programme schedule. But why stop there? You can use channel and programming affinity to estimate congruence to the target using scores, Pearson coefficients or other statistical functions. The join to RPD is on the IP address.
Advertisers speak to an individual, not a household. Most purchasing decisions are made by individuals, not households. Besides, households can differ markedly as well (this is hardly a revelation). Imagine two adjacent households (the customer database has the addresses): similar income, level of educational attainment… In one household, soft drinks are consumed voraciously and brand-loyally; the people who live in the other house, however, disdain them and drink mineral water, filtered tap water, and tea. One household has two Toyotas for the reliability and reasonable price. The other is loyal to Volvo for the safety and aesthetics, and because they want to support the company’s working ethos, which includes a reliable social safety net and eschews conveyor belts. The first household has plenty of young children and buys their sundry supplies by the truckload; the other family’s children are teenagers or young adults and they need none. The two households also vacation, dine out, travel and consume music differently. From almost any advertiser’s perspective, they couldn’t be more different. Yet the pay-TV operator’s CRM, which stores their next-door addresses, considers them identical. This insufficiency of demographic and consumption descriptors is what’s holding many pay-TV operators back from making high CPMs in advertising.
B. With personall, you see your subscribing household members’ use of other services, such as your and competing OTTs. You can empirically determine the OTT share of viewing threshold beyond which intervention makes sense.
C. Come fully equipped to the negotiations over retransmission and content. If you have RPD, you already have an idea of the quantity of viewing to channels on your system and some notion of the success of otherwise licensed programming. But this doesn’t directly answer the question: how many subscribing households would I lose if I didn’t renew the contract with channel XYZ? To determine that, you need to know how central is this channel to the utility of your individual service to each viewer in the household. personall tells you. For example, if a teenager watches channel XYZ 10% of his overall viewing time on the system (indisputably high), but doesn’t use your system much and gets most of his video from Netflix and Amazon Prime Video, he is unlikely to make a big deal of your nonrenewal of XYZ. On the other hand, someone else in the household, of age 65, watches XYZ 5% of her system time and her nonsystem time is zero (she never uses OTT at all). If you take away her XYZ you’ll quite likely lose her to the IPTV service from one of the mobile operators, which promises to make the setup look exactly like cable, and which has no intention of dumping XYZ. And then you have the person whose name is on the bill. She uses XYZ for merely 2% of her system viewing, and watches a lot of Netflix and some Amazon Prime Video besides. But you touch her XYZ and you’ve lost her. It’s the principle of the thing.
A. What if you are a manager of one of the OTT services (most likely from a telco) instead? Your system has individual user accounts, and you even capture some individual viewing from your app on mobile devices. Well, yes and no. Yes in that you are capturing some individual viewing. No in that much of the viewing you’re capturing is not necessarily individual. Most households still contain one, and often more than one, big screen, which offers a different comfort vs. privacy and mobility tradeoff. Do you know whether the authorised users of the account log in and out of their user profiles as you would have them do? What if they leave the app on the connected TV logged in on the same user profile, to minimise effort and get the benefit of autocontinuing playback separate from their personal devices, on which they watch other content? Of course you are concerned by the revenue impact of password sharing, but you understand that overtightening the screws on “TV everywhere” would undermine one of the foundational ideas of its existence. So you don’t do that. And that leaves you with unreliable user profiles, and thus in exactly the same bind as pay-TV operators: doubtful individual attribution and no demographics other than the address.
You haven’t been able to rely on the currency services because their samples are too small to be relatable to your subscribing household base, and because they cannot measure parts of your operation (VOD, OTT, mobile). But now you can rely on someone. This time, it’s personall.
personall / Alldience is ready
15 April 2020
Our state-of-the-art audiovisual audience measurement system, personall, available integrated with eCGlobal Research Solutions’ Latin American panels and advertiser services as Alldience, has passed rigorous testing and is ready for deployment anywhere.
personall / Alldience can measure any content with sound played on any device and any platform, classic TV, DVR, VOD, OTT, Web, app, social media, outdoors, fixed or mobile.
personall / Alldience has specific applications for broadcasters and pay-TV channel operators; advertisers, marketers and agencies; pay-TV system operators/MVPDs (enabling advertisers to target your advertising inventory on an individual, demographic basis, and you to prevent churn); OTT system operators; public broadcasters; and radio.
Our outstanding quality, based on 33 years of experience in audience measurement system development in combination with one of the world’s finest ACR systems, is also among the most affordable solutions.
As advertising and marketing budgets are necessarily reduced in the aftermath of the Covid-19 crisis, can we help you?
personall / Alldience: the complete audience measurement system that measures any screen, any source, anywhere, anytime completes a six-month pilot deployment in Brazil
17 May 2019
Immetrica, audience-measurement systems engineering specialist, and eCGlobal Research Solutions, innovative consumer insights and big data analytics company, are proud to introduce personall, offered in combination with eCGlobal’s Latin American panels as Alldience, a new ratings service built from the ground up to capture all viewing regardless of location and technology, and integrated with eCGlobal’s existing social-media use and consumer panel with approximately one million active members served by dozens of pay-TV operators. The service can measure new-technology-enabled viewing to realtime and timeshifted DVR but also sources invisible to most currency ratings providers, including VOD, OTT, websites, apps, YouTube, other social media, in and out of the home. Based on a highly reliable smartphone-enabled ACR technology, the service detects viewing in any environment in which the media audio remains intelligible, in realtime or timeshifted using any means for up to a year after broadcast—or longer.
For advertisers, we offer complete audience data that permits the calculation of campaign ROI; campaign retargeting that permits the purchase of inventory that reaches the target audience; ad creative testing, covering propensity to watch in realtime and on DVR, opinion and biometrically measured emotion; and reporting on the entire consumer journey.
For premium-channel operators, we offer channel-coverage ratings (ratings among their subscribers rather than the entire sample) when eCGlobal operates subsamples of its panel specific to them. We can also offer a behavioural definition of channel use, which, however, ignores viewers who have access to these channels but do not use them.
personall can measure binge-watching, and can ingest the multinational streamers quickly and efficiently, and update their libraries promptly for additions and deletions. personall can support near-real-time programmatic advertising markets and other uses that require near-immediate turnaround of data.
The unit of measurement is a demographically described individual viewer, with a rich single-source set of descriptors such as income, education and occupation; social-media use and engagement; and consumer behaviour available to drill down further to the target or de facto audience, on a panregional, nationwide, or local-market scale. In addition to audience measurement for broadcasters, we offer advertising spot ratings and competitive analysis of advertising flights.
Of particular interest to advertisers is the opportunity to engage panel members who have just watched designated content with a custom survey automatically pushed to the members’ smartphones in near-realtime, in what amounts to the world’s largest focus group.
The service is delivered through a world-leading analytics platform capable of handling the large sample sizes; an optional cost-effective custom reporting service using Immetrica’s proprietary technology; and customized means for clients with needs beyond these.
A test deployment in Brazil returned first usable data on 31 August 2018 and continues to run today, demonstrating high accuracy, stability and utility. We have also studied the viewing of 84 designated advertising creatives.
Contact us to discover what we can do to find all your audience.
More on personall »
Immetrica and eCGlobal introduce a comprehensive, accurate, affordable ratings service
2 February 2018
eCGlobal Research Solutions, innovative consumer insights and big data analytics company, and Immetrica, audience-measurement systems engineering specialist, today announce personall, a new audience measurement system. In Latin America, personall is based on eCGlobal’s existing social media engagement and consumer panel with half a million active members served by dozens of pay-TV operators, and branded Alldience. The service will have by far the largest sample sizes available in any country in the region, bringing unparalleled accuracy to today’s multichannel realtime and timeshifted DVR / VOD / OTT media use. Based on a highly reliable smartphone-enabled ACR technology, the service will detect viewing in any environment in which the media audio remains intelligible, in realtime or timeshifted using any means for up to a year after broadcast.
For premium-channel operators, we offer channel-coverage ratings (ratings among their subscribers rather than the entire sample) when eCGlobal operates subsamples of its panel specific to them. We can also offer a behavioural definition of channel use, which, however, ignores viewers who have access to these channels but do not use them.
The unit of measurement will be a demographically described individual viewer, with a rich single-source set of descriptors such as income, education and occupation; social-media use and engagement; and consumer behaviour available to drill down further to the target or de facto audience, on a panregional, nationwide, or local-market scale. In addition to audience measurement for broadcasters, we plan to offer advertising spot ratings and competitive analysis of advertising flights.
Of particular interest to advertisers is the opportunity to engage panel members who have just watched designated content with a custom survey automatically pushed to the members’ smartphones in near-realtime, in what amounts to the world’s largest focus group.
The service will be delivered through a world-leading analytics platform capable of handling the large sample sizes; an optional cost-effective custom reporting service using Immetrica’s proprietary technology; and customized means for clients with needs beyond these.
Launch of clean-sheet RPD measurement system
1 October 2016
We have completed an entire measurement system for a multinational client, a clean-sheet design.
Features (a partial list):
  • Empirically determined capping
  • Industry-standard intab definition
  • Support for two dissimilar data collection systems
  • Channel coverage ratings—a regional exclusive
  • Adjustment for SD/HD imbalance in sample relative to universe—first in the world
  • Extrapolation of STBs to households, in the absence of whole-household measurement of many households—first in the world
  • Integrated, automated sample management
  • Maximally automated channel management (complete automation impossible as the two operators supplying data have no stable channel identifiers)
  • Operations monitoring and alerting of exceptional conditions
  • High-efficiency parallel processing
Development roadmap:
  • Time-shifted data service (underlying technology already implemented)
  • VOD
  • Advertising spot ratings
  • Sample balancing by demographics, geographics, psychographics (in this region, geographically matched income/education/occupation data) and technical parameters
  • Local-market ratings
  • OTT
  • Personalisation of STB data (its relation to individual viewers) by matching expected data from Netflix-like accounts for use of broadband services
During development, there were several showstoppers—bugs or spec deviations in ancillary systems or STB middleware, each of which would have stopped or seriously degraded the service. Audiresys discovered these problems and participated in the effort to correct them.
Another phase of a comprehensive audit
21 January 2018
We complete another phase of a comprehensive RPD audit for an Asian client with a majority market share in its country—but facing challenges.
The audit covered STB management (deactivated STBs that returned data and active STBs that did not), conditional-access issues, inconsistencies in published data, regression testing of the data collection software on all STB models, intab formulation including design and construction of a benchmark proof-of-concept computation system, a much greater rate of exclusion from the intab on two similar new STB models relative to older ones from other manufacturers, several stability studies after data collection software changes, channel management, schedule comprehensiveness, and many other issues, revealing and correcting numerous problems. A key achievement was a negotiated package of changes to the data collection software that qualified it for use, including the addition of information used in calculating the intab. Work is ongoing.