S4C is tendering for the supply of viewing and audience appreciation data services in relation to S4C programmes. The tender has been advertised on the Sell2Wales.co.uk website and in the Official Journal of the European Union. All businesses interested in providing the services must complete a Pre-Qualification Questionnaire in the form set out below. The deadline for receipt by S4C of the completed Pre-Qualification Questionnaire is midday on Monday 9th of February 2015.
All requests for further information and/or guidance in completing this PQQ and/or this tender process should be addressed to: firstname.lastname@example.org Please note that no queries or requests for information relating to this tender process should be addressed to any individual officer, employee or representative of S4C.
The closing date for receipt of requests for additional information and for guidance in completing the PQQ is midday on Monday 26th of January 2015.
Section 3, part b) – Is there a limit on the number of examples of relevant contracts you would like to know about or should we include details of any contract we feel relevant (in the past 5 years)?
There is no limit. You are asked to include any examples over the last 5 years that you feel are relevant.
Please can you provide more details about the required panel composition:
Representative of Wales overall or representative but with specific segments of interest (e.g. regular S4C viewers?)
In terms of the Welsh-speaking panel members, how are they defined? I.e. are they defined as people use Welsh as their first language, or as people who have the ability to communicate in Welsh to some extent and therefore watch S4C?
We would want the panel to be representative of Wales and Welsh speakers as noted, but we could also filter the results by viewership of S4C to create an additional reporting group of "regular S4C viewers", however defined.
The panel balance is obviously over representative of Welsh speakers, these need to be weighted to the correct proportion when reporting "Wales".
We would be happy to consider one of the two options below.
We currently use the following graded fluency levels (we have BARB universes on these fluency levels).
FluencyLevel 1 Understand, speak, write and read Welsh EXTREMELY WELLLevel 2 Understand, speak, write and read Welsh QUITE WELLLevel 3 Understand, speak, write and read A LITTLE WelshLevel 4 Can understand and speak SOME Welsh
We have previously used and other public bodies also use the following fluency levels: fluent; can speak a fair amount; can speak a little Welsh; or can only say a few words. (We have universes for these).
Do you currently have a panel – and if so how long has it been running?
The current panel has been running for 7 years, and similar services were conducted before this period by other companies.
Will the surveys be mainly the same each "week" or will there be regular / ad-hoc changes to the questionnaire?
You will be receiving S4C's TV schedule each week for the Viewing Diaries, (so the survey component for viewing & appreciation of programmes is mainly based on that) with a few common short questions each week – no more than 2 or 3.
You may want to consider how you would like to receive these – we can send them directly (with updates/changes) or you may want to receive from a source such as BDS (Broadcast Data Services). Currently both sources are used, (first draft from BDS, with late changes from the S4C source).
Also, around once a month we will be asking ad-hoc questions about around 3 topics in depth, to vary each month.
For each survey, is there a minimum base size required for any specific groups?
We tend to report on AI's monthly with AI samples of 25 or over within each genre. We also, however, need the ability to see smaller sample sizes.
We need to see sample size for all reports so that we can gauge reliability. Please advise if you believe that minimum base sizes should be used.
Can we have an example of a similar questionnaire from previous research initiatives?
We will not be sharing a questionnaire – they are very much industry standard. They have 2 elements – a viewing diary with with a 10 point appreciation scale against programmes and questions about method of viewership for each programme, and a further classic survey element with a combination of closed and open-ended questions.
Do they have any estimated response rates for weekly and/or monthly surveys?
We would rather ask you for estimated response rates based on your own experiences of such panels.
Do you require any other deliverables apart from access to results? E.g. any reporting, presentations etc.
The deliverables depend on your delivery method. If we have access to software to report on everything, we would not specifically expect any physical reports or presentations, though please offer these if you believe that they may enhance your offering.
How many face-to-face meetings do S4C envisage as required in addition to the annual meetings between them and the successful applicant?
We do not envisage the need for any face-to-face meetings other than annual ones as a matter of course, though possibly in the first year we may want to meet quarterly. We also envisage conference calls by telephone or video-link, as needed.
Please can you clarify the maximum length of submission – is it a maximum of 15 sides of A4, or 15 pages of A4 (i.e. 30 sides)?
It is 15 sides of A4.
We are assuming that all questions will be circulated to all tenderers – please can you confirm that this is the case
The questions and answer will all appear on the S4C tender website, and all tenderers will be directed towards it.
The ITT mentions two closing dates for receipt of questions in connection with the tender, could you please confirm the closing date for receipt of questions?
All questions must be received by S4C by 12.00 on 9th March 2015.
There will be an on-going weekly survey based on individual S4C shows (52 waves across the year). For this element do you need to understand day of week viewed i.e. diary format or will it simply be enough to know they were viewed that week and length of time they were viewed for. This would be alongside the AI score collected for the show. Also do you have any idea of the number of shows you would want to ask about each wave?
It will be enough to know they were viewed that week and length of time they were viewed for (and viewing method). In each week S4C will be asking about up to 60 S4C programmes per week. Please note that the viewing diary will ask about S4C programmes only.
For the monthly task you have requested channel appreciation scores – would these be aggregated up from the scores of individual shows or just taken as a single question asked at the channel brand level?
The viewing diary will be S4C programmes only, so S4C AI's can be derived from those. We also intend asking channel-brand level questions monthly (and S4C will be among that list too). So in effect we will have two methods analysing S4C's appreciation.
Also if all channel scores are to be based on aggregated data taken from individual shows would your expectation be that the weekly S4C shows survey be merged into the monthly survey for those waves or would it remain as a separate survey each time?
S4C's AI scores would need to be available separately by week, but also merged for monthly (or any other period) of reporting.
If we win the project, will the existing panel be migrated to us from the current supplier? If so, will be entire panel be migrated or will we need to invite them to join the new panel rather than the incumbent's version that they've always been on.
The current thinking is that the existing panel will not be migrated.
I just wanted to clarify the demography of the panel, given that 50% of the panel need to be Welsh speakers while only 20% of the population of Wales speak Welsh. Are we looking to recruit half the panel that fits the demography of Welsh speakers and then half to fit the demography of non-Welsh speakers?
Yes, we looking to recruit half the panel that fits the demography of Welsh speakers and then half to fit the demography of non-Welsh speakers, then to weight to total population when reporting on that Total category.
Is 1,500 the size of the gross panel or the size of the reporting panel?
It is the size of the gross panel.
Does the panel currently incorporate a portal website into its operations for respondents to visit and refer to, or is the panel treated more like a database where everything is done via our correspondence with them?
A website is currently used but that's mainly for Q&A type information for online panellists in order to reduce email volume. The panel is mainly treated like a database where everything is done via correspondence.
Could you tell us how the panel is currently incentivised? If we are looking to migrate existing members across, I wouldn't want to suggest anything in our proposal that is wildly financially different to what is used now. If we offer too much, we'd be using up budget on something that would be better spent elsewhere (such as recruiting more members) and if we offer too little, it won't be any kind of enticement for existing members to switch.
The current fixed incentives are a total prize draw fund pot of £700 a month, plus a fixed £5 a month in high-street vouchers for all 16-34s who complete a survey online that wave. Additional prizes during the year are also added to help maintain engagement so items like games consoles or tablets normally linked to an event like a Halloween special, but these are not fixed in any way.
The brief mentions that 70% of the surveys should be online and at least 30% offline (post). These proportions are not mentioned in the brief in terms of the make-up of the panel itself, so can we assume that 30% of our panel should be offline and recruited as such?
This refers to panel composition and not response rates. So yes, 30% of the recruited panel should be offline.
Given that online research and postal research work at very different speeds, what kind of timetable have you become accustomed to in terms of fieldwork lengths and delivery of reporting?
We'd expect everything to be delivered approximately 4-5 weeks after the diaries/diagnostics were distributed to respondents.
A follow on to Q21 & Q22 – given that online research and postal research can generate very different response rate levels, I'd also like to know what response rates you have become accustomed to over time on the panel. I hope you can let us know, both for online and offline respondents. In particular I'd like to know this so that we can make calculations on how what proportions of online/offline members we should look to recruit in order to achieve at least 30% offline responses. If you are used to offline generating a much larger response rate than online, for example, then we'd look to balance our up-front panel recruitment accordingly.
Please note that the 70/30 split refers to panel composition and not the composition of responses. We will not be providing response rates, we would rather ask you for estimated response rates based on your own experiences of such panels.