MarketTools Blog

This is the third installment in an ongoing series looking at the invaluable list of 26 Questions to Help Research Buyers of Online Samples assembled by ESOMAR, the global non-profit market research organization.  You can review the other entries in the series here.

This time around we’ll explore the policies of panel companies for blending sample: “Do you supplement your samples with samples from other providers? How do you select these partners? Is it your policy to notify a client in advance when using a third party provider? Do you de-duplicate the sample when using multiple sample providers?” (Question #20 on the list.)

Ultimately I think it’s fair to say that most panel companies have at some time engaged in the practice of blending their samples with those of other sample providers (or will in the future).  Reading between the lines, you might realize that while there are four separate questions above, they really boil down to a single question: “Is this ok?”  The answer, in my opinion, is “Yes, it can be,” but I’m getting ahead of myself.  Let’s tackle each question in turn and dig into the key issues related to each and I’ll show you how I reached my conclusion.

1.    Do you supplement your samples with samples from other providers?
Make sure you get a clear “yes” or “no” response – anything else is already headed in the wrong direction.  There are a couple of key reasons why you might want your panel provider to blend sample:
-  You need them to deliver sample from countries where they don’t operate their own panels.
-  You are conducting studies that require low incidence respondents that really cannot be delivered from a single panel.

This should be viewed as a value-added service that a panel provider is engaging in on behalf of their clients.  Many clients do not have the procurement infrastructure to use multiple providers to fulfill their sample needs, or do not have the volume of business to engage with multiple vendors and still keep their operations efficient and cost-effective.  Where low incidence studies are concerned, the ability to blend sample can be viewed as an insurance policy that gives clients the confidence that their survey can still be completed even if there are changes in scope or incidence while fielding the study.

2.    How do you select these partners?
This is a much more critical question, and starts to get into the more important aspects of blending sample – the practice does require that sample providers be highly selective in which partners they work with and what they require of them.  At a minimum, look for a firm that conducts extensive reviews of the industry (50+ sample vendors) to both stay on top of best practices, and to make sure that they select only the best partners to blend their sample with.  In addition, they need to implement and mandate participation in an industry-leading data quality validation program (such as TrueSample).  With these requirements in place, clients can have confidence that their providers have their best interests in mind.

3.    Is it your policy to notify a client in advance when using a third party provider?
Transparency is a key issue in this situation, and the answer again should be a clear yes or no.  Ideally a panel provider and their clients will work diligently to establish a candid and honest relationship.  Ultimately, the provider’s responsibility is to make sure that the research they are delivering to an end client is of the highest quality and as accurate as humanly possible.  Our industry does not have a place for antiquated ideas of secrecy.

4.    Do you de-duplicate the sample when using multiple sample providers?
Yes, and only yes, is the appropriate answer to this question.  The technology exists to de-duplicate sample, and it’s a critical process that relates to the quality of the data being supplied.  Panel and sample suppliers must have proactive and clear policies on how they identify and manage overlap across panels as well as analyze panelist behavior within a survey.

In conclusion, blending sample with other providers is an acceptable practice, and it's also a value-added service that drives real value to clients – provided that the sample supplier has taken the proper steps to identify and select the best supplemental sources, and implements an automated technology-driven data quality solution.
 

As customer loyalty researchers, we recognize that at the root of every loyal or disloyal customer is a history of interactions—calls with sales, tech support, maybe an installer or a billing specialist.  We know that each one of these customer interactions has the power to impact customer loyalty.

The survey process is an interaction as well.  It is a chance for the customer to tell you how well you are performing.  Like any other interaction, it can be handled well or handled poorly.

Our best advice to researchers designing their feedback program is to channel the principal tenet of the Hippocratic Oath, “Do No Harm”.  This should be the primary consideration in the decision making process at every step of designing/implementing a satisfaction measurement program.

Obviously the first step is to work with the right partner so you know you’ve designed an excellent survey (just the right length, focused questions, only the required open-ended questions, etc.).  This is critical.  But don’t spend weeks refining and vetting the survey instrument, just to leave other key parts of the interaction to the last minute (or even until after the survey is live!).  Here are some other areas that should be given equal consideration when designing a feedback program.

Pre-Survey:
A personalized email is generally preferred to a generic greeting.  However, personalization requires a reliable database.  I would rather receive an email with “Dear Valued Customer” than any of the following:

Dear MIKE MILBURN  (why are you yelling at me?)
Dear mike milburn  (only acceptable if gauging opinion on the poetry of ee cummings)
Dear milburn mike  (do I even need to explain?)
Dear tbd unknown  (am I a client of yours or not?)
Dear Mrs. Milburn  (would be ok if it was for my wife, this one wasn’t)
(By the way, avoid Mr., Ms., or Mrs. unless you are absolutely sure, and even then, don’t)

Is your database ready for the survey process?  Even if it is accurate 99.5% of the time, 10,000 emails could yield 50 upset/turned off invitees.

Follow up on specific requests:
Often companies will include a survey question asking respondents if they want someone from the company to contact them.  If a respondent says yes, then contact them — you have now set an expectation that you need to follow through with.  It sounds simple, but many companies turn this kind of opportunity to interact with an individual respondent into a reason for the respondents to feel as if the company doesn’t value them when they neglect to follow up.  Loyalty is an emotion; hurt feelings will quickly damage the relationship.

Also, companies may ask if the respondent needs certain materials, documentation, etc.  Again, if someone says they  need it, provide it.  I have seen programs where, 6 months in, the client has removed this type of question because they realized no one in their company had been following up on it since the program began.  That is 6 months of setting expectations and not fulfilling any of them.

Do you have the infrastructure in place to deal with the volume of contact or information requests?  If not, you are better off not asking these questions until you do.

Survey follow-up (in general):
The ability to follow up in real time with respondents who give low scores to your survey can be very powerful.  Imagine this from the respondent’s point of view:  Let's say I had a bad experience, I expressed it in your survey and that same day I get a phone call from your company.

“Hello, is this Mike Milburn?  Thank you for providing feedback recently on your service experience.  I understand things were not handled up to our usual standards.  I’d like to take a minute to discuss…”

Think about it.  I am no longer a nameless person throwing my feedback into a void.  Someone saw my responses and acted on it.  I’m now part of the solution.

Much better than this call:  “Is this Mike?  Yeah, Mike, why did you give us a 2 on courtesy and professionalism?”  (Hmm, I wonder.)

Companies spend millions of dollars putting the right messaging campaigns together.  Yet, when following up with respondents in critical one-on-one interactions, these events are often left to the whims of the individual reaching out to the client.

Do you have a scripted, structured follow-up mechanism in place?  If not, how do you know these interactions are being handled in a consistent, professional manner?

A feedback program is another opportunity to interact with your clients.  How you interact with them is up to you.  At every step of the process, let “Do No Harm” guide your decisions.

We’ve been hearing more and more about how U.S. patients are taking control of their healthcare.  We also know that patients see and recall pharmaceutical advertising, and in some cases ask their physician about a particular product. In fact, a recent study from Prevention Magazine noted that a majority of consumers feel that pharmaceutical advertising in magazines and on TV is fair and balanced, but they feel that online ads need to work harder to communicate risks vs. benefits.  At the same time, advertising within the social media landscape has quickly become an accepted voice for pharmaceutical companies, with 57% of consumers saying ads are acceptable on sites that cover health and medical issues.  Good to know!

MarketTools ran some market research on research recently with a study of over 1,200 consumers, and found some further validation of the ways that both general consumers and specific patient groups continue to seek information and tools online related to their healthcare.

• Web MD is considered an important resource by nearly 50% of consumers overall, and by over one-half of consumers with specific diseases.
• Nearly one-half of consumers and patients would consider a doctor’s website a good place to go to learn about their disease.
• Interested consumers also consider offline medical textbooks a good place for learning about diseases.
• Most consumers in this study don’t consider TV, newspapers, pharmacies or quick clinics good places to learn about their disease.  
• Blogs are increasingly becoming a source of advice for patients:  11% of consumers trust a patient blog more than information from a pharmaceutical manufacturer website. That percentage jumps slightly when you look at patients with chronic conditions, such as chronic pain.
• Nearly 11% of consumers and patients interested in information about chronic pain, diabetes and heart disease participate in social networking.  
• And confirming what many in pharma already know: only about a quarter of consumers or patients would consider a pharmaceutical website a good place to learn about their disease.

We’ve long known the Internet is a great tool for massive patient education, and clearly, pharmaceutical marketers are continuing to invest in their online presence including websites and ads.  But it seems like this current research study spotlights a great opportunity for pharmaceutical companies to assist doctors in beefing up their web presence by offering them informational content and giving them opportunities to connect with other medical professionals (such as nurses) who can help educate them about products.  There are additional opportunities for pharma to get involved in the education process and identify new resources for doctors and patients to find trustworthy content.  For example, we anticipate that social media participation around health and medical issues will increase transparency and  communication between patients.  We encourage our pharmaceutical clients to optimize the best combination of online and offline media to result in the greatest reach to their target market.

Coming up soon:  study results about what patients want in terms of online tools and technology to connect with healthcare information, including how they feel about the possibility of online test results and medical records. Stay tuned!
 

One of the things I enjoy most about working at MarketTools is the ready access we have to true experts in the fields of survey design, market research, customer loyalty, survey panel management – you name it – who can provide best practices advice. Just last week I was fortunate to interview Jolinda Decad, one of our MarketTools CustomerSat Research Consultants, to discuss survey design best practices for EFM (enterprise feedback management). Here are some highlights of what I learned – or you can hear the entire interview here (about 10 minutes).

•  Focus, Focus, Focus:  As you begin, resist the urge to pull together a cross-functional team to start brainstorming survey questions. The best surveys focus on a specific purpose and ask only questions that are relevant and actionable.  The more people involved, the more likely the purpose will be clouded and the more difficult it will be to get the information you need from the survey.
• Clarify survey objectives:  What problems are you trying to solve by asking customers about their experience? Often companies are looking to understand how to meet their customers’ evolving needs within a particular touchpoint, and it’s helpful to think through the end-to-end customer experience within that touchpoint.  For example, if you’re evaluating customer satisfaction in your contact center, you’ll want to think through the typical contact center experience and determine questions you could ask about each aspect of the experience, such as how long they are on hold, how they are greeted, how their issue was handled, if there was any follow-up, etc.  
• Anticipate Action:  Think through how you might incorporate the survey results into decisions and actions. This will help you in two ways:
  - First, it helps you determine the demographic variables you need for decision-making.  If you want to make decisions based on customer segment, region, product lines, etc., you can ensure you include these demographic variables in the survey to more easily segment the data for analysis.

  - Secondly, you can ensure that questions are asked in a way that drives clear action without setting false expectations. For example, if you sense that some customers want longer support hours but need a better idea of the number of customers that actually require this, you will want to ask whether the current support hours are meeting their needs without asking a question like “would you use support after-hours if it were available?” Not only is this question leading the respondent, it also sets the expectation that you’re considering extending support hours, which you may determine isn’t necessary if only a small percentage of customers feel their needs are not being met.

That’s only the beginning – learn more about how to design a great customer feedback survey from one of our EFM best practice consultants.
 

From time to time, I hear industry pundits talk about how ensuring research quality is as simple as applying a bunch of disparate tools throughout the research supply chain. These folks suggest that panel companies could match their panelists against industry databases to cleanse out non-real panelists, and that research companies can prevent duplicate panelists and speedsters by applying digital fingerprinting and proprietary speeding measures within surveys.

In my opinion, piecemeal solutions like these don’t address the most critical aspects of quality that clients have been pleading for: “transparency” and “consistency”. If a buyer has no way to audit or visually examine the overall impact that each quality assurance tool has made on their research, then they have no way to measure the quality of the project or the supplier. This doesn’t seem like a fair trade-off for the clients who have stated outright that they are willing to pay a premium for quality – but only if they can measure it and depend on it.

If you haven’t already, I highly recommend reading David Haynes’ article on “Lemon Markets”, which was featured in Quirk’s Marketing Research Review in August, 2010 (subscription required).  As the Chairman and CEO of sample supplier Western Wats, David skillfully makes the point that the online sample market has been acting as a “lemon market” in which suppliers cannot visibly demonstrate quality before the sale, and therefore, buyers can’t evaluate quality either. The common consequence of a lemon market is that suppliers are incented to reduce quality – because without any way of measuring or determining quality, buyers won’t be willing to pay more for it. David says that in these cases, “Objective third party quality validation can augment reputation and enable sellers to credibly demonstrate quality before the sale.” He goes on to say that “Third party quality validation must provide relevant, unbiased, standardized metrics comparable across sample suppliers.”

I would argue that piecemeal solutions where various technologies are applied by different parties in the supply chain do not visibly and credibly demonstrate quality. Only a comprehensive, auditable and consistently applicable solution, like TrueSample, achieves this goal.

Clients like Procter & Gamble and Microsoft are beginning to require that a standardized quality solution be applied to all research projects across all vendors and sample suppliers they work with. And more than that, they want real-time reports that show them exactly how many respondents were excluded and why. This consistent transparency into the "black box" that is online research quality is what TrueSample provides.
 

One of the questions clients often ask me is “how do my NPS scores compare to others?” (For those unfamiliar with the term, NPS, or Net Promoter Score, is a metric for measuring customer loyalty.  NPS is the percentage of  “Promoters”, or active, loyal customers likely to recommend your company, minus the percentage of “Detractors” who may spread negative word of mouth.) 

While NPS benchmarks by industry and best-in-class scores exist, there are really no better scores to benchmark yourself against than your own.  Really, the most important thing to remember is that you are looking for improvement.  When thinking about comparing your scores with other companies, please remember that there are many things that can affect your NPS score – including the industry you’re in, the type of products you sell, where you sell your products, who you sell your products to, the economy, etc. 

Also, remember that NPS is only a single metric, one of many that should be tracked and monitored to measure performance.  NPS should not be relied upon alone, any more than a pilot would rely on a single gauge in the cockpit to monitor his flight.  Use NPS in conjunction with overall customer satisfaction scores and other actionable measures (like detailed performance metrics, individual customers’ scores, etc.) to garner the greatest insights into your performance.

Based upon conversations with clients using NPS, I would suggest the following strategy for benchmarking NPS:

1)    Establish your own benchmark

• The only benchmark that is applicable for you is your own. Measure for a time period (a quarter, a year) and then set the average as your baseline.
• Set a goal for increasing your NPS score year over year. The important thing here is to improve.  Do not get caught up in determining by how much.  How much depends on several factors that are usually out of your control – budget, ability to make changes based on data, resources to affect change, internal company culture, economy, etc.
• Year over year, or rolling 6-12 months is the best way to trend, as you will most likely see dramatic changes in scores from month to month or quarter to quarter.  See point #2 below.

2)    Watch your mix

• Develop a sample plan that allows you to keep your survey population mix as comparable as possible from time period to time period.  For example, if 75% of the people you survey in  Q1 are in North America and in only 40% of the people you survey in Q2 are in North America, you will see dramatic differences in your trending.
• Typical things to think about when developing your NPS survey sample plan include:
  -    What type of people you are surveying?  Titles, positions, and other factors affect survey scores.
  -    Where are the people are located? Certain countries will typically not give high scores of 9 or 10 out of 10 when responding to surveys, so is NPS the right metric when measuring that region?  Whatever you decide, the mix of countries should be comparable.
  -    How long have your survey respondents been clients? If your Q1 sample is 75% customers that have been with you 3 years or more and your Q2 sample contains only 40% of customers that have been with you 3 years or more, this could dramatically affect your trending.
  -    Where in the survey are you asking the question? If you ask it at the beginning, you will get a higher response to that question, but only a “top of mind” answer.  If you ask at the end, you will get fewer responses to this question, but the responses you get will be more thought out after answering all the other questions.
  -    What languages did you use for the survey? If the survey was written in English and fielded to people whose primary language is not English, they may not truly understand the question.  When asking in other languages, it is imperative that the translation is done in a way that explains the question properly.

These are just some of the many things to think about when using NPS as a metric, but perhaps most importantly, don’t lose sight of the fact that there are many other metrics out there that may be more useful to your organization – stay tuned for further discussion on other useful metrics to consider.
 

Since its launch a few months ago, the iPad has already had a positive impact on market research in both the consumer engagement and data collection aspects of fieldwork.  USA Today noted that Apple’s new tablets are luring in curious consumers to participate in research projects just for the chance to try one out.  The iPad offers field researchers two key benefits:
1.    The data collection process automates what was once manual with paper surveys
2.    The intuitive touch screen interface of the device is proving to engage potential respondents more effectively than pen and paper

With over 3,000,000 iPads sold to date and similar devices like the HP Slate set to be released in the coming months, the touch screen interface is rapidly becoming mainstream.  A study looking at the “iPad effect” has also noted that owners are using iPads in more ways than anticipated – in place of specialized gadgets like e-readers, gaming devices, laptops, and smartphones.  It’s important, then, to consider the impact these new devices and their engaging interfaces may have on online market research and online panels.

Online panel companies are constantly working to increase the number of active survey takers on panels – in general, by improving the respondent experience before and after the actual survey through recruitment and incentive programs.  The growing popularity of the iPad presents a unique opportunity to deliver a new and enhanced respondent experience during the survey.  For example:

•    The touch screen technology has panelists making a direct, physical connection with a survey vs. the more abstract point-and-click approach that is prevalent today.  In general, the question and answer format of surveys hasn’t changed dramatically since the transition from pen and paper to online surveys.  The idea of using gestures to select, drag and drop could create entirely new survey environments that are much more engaging to the panelist – like selecting items from a virtual grocery shelf and putting them into a virtual shopping cart.  The more engaged the panelist, the more likely they are to come back for another survey.  
•    In turn, panelists will be more easily able to process responses with their fingertips vs. with the mouse, most likely leading to the ability to complete a survey in less time.  Even if the actual survey lengths do not decrease, it may feel shorter to respondents if the survey process is more fun for them – leading to an improved survey experience.

The iPad could make the online survey-taking experience more attractive by using the new touch screen medium to full advantage, opening up opportunities for research companies to tap into a broad pool of new potential survey takers.  This could increase the volume of panelists in the U.S. both by giving current panelists an interest in extending the amount of time they spend as active survey-takers, and by providing an engaging, rewarding experience to a population of new panelists that may have never considered participating on a panel before.
 

Geographic portability, or the ability to apply a technology used for market research across a variety of countries, has been identified by Microsoft and Procter & Gamble as one of their selection criteria for online research quality solutions (noted by my colleague Emily Morris in her recent blog post, and by Reineke Reitsma of Forrester).  Research companies who can offer their clients geographic portability in their solutions to validate survey respondents and panelists are able to ensure better data quality, as well as reduce costs for global market research and allow for easier comparison of results across countries.

However, few data quality solution providers are able to offer geographic portability for a variety of reasons.  One major reason is differing privacy laws and other regulations across countries.  For example, digital fingerprinting technology is very commonly used in the United States to identify when an individual panelist attempts to complete the same survey twice, but may be problematic in other countries, depending upon whether a particular implementation collects too much or the wrong types of data.  The Marketing Research and Intelligence Association (MRIA) has warned its members that digital fingerprinting may be unlawful in Canada as it “may collect more information than is necessary to identify fraudulent and duplicate respondents in online research”.  Canada’s comprehensive privacy law is similar to data protection laws found in Europe, Argentina, Australia, New Zealand and Japan – which means that solution providers need to be careful to ensure their digital fingerprinting methods conform to local privacy regulations.

Another challenge is identity verification.  In more mature markets (such as the United States and the United Kingdom), access to identity verification data is relatively easy – so a data quality solution like TrueSample can use name, address, and other data to verify survey respondents and panelists.  In less mature markets, this kind of data has not traditionally been available – but a number of trends are improving the situation.  For example, as China has grown and developed, the establishment of state and private credit bureaus and the emergence of data providers like ChinaLOOP (now Axciom China) provide ways to verify that survey respondents and panelists are who they say they are.

More and more multinational businesses are conducting research projects that cross international boundaries – which means that the need for online research quality solutions that meet geographic portability requirements will continue to grow.  Currently, the TrueSample data quality solution is one of the few solutions that meets geographic portability requirements across the US, Canada, Germany, France and the UK.  As MarketTools and other companies develop and enhance technologies to expand geographic portability, market research clients will gain greater trust that all researchers are meeting an expected level of quality – even when the survey sample extends across the globe.
 

For years, people have told me I should write a book.  I ask them what should I write about and they say, “You’re always telling stories, why don’t you write a book of all of your stories?”  Then I ask, “Well who would read that, let alone buy it?”  And they say, “Well, I would.”  I guess that this is a rudimentary form of marketing research similar to “Do you like my new dress?”  You can ask a question and get an answer, but at the end of the day, this is misleading research which will lead to a bad business decision.

Now - back to the point about me writing a book.  I’m not sure that I have a book in me, but I sure do have a lot of chapters that I could write.  And the first chapter is What Are We Trying To Do?

When I was on the client side, the Marketing teams would come to me and say, “We need an Attitude and Usage Study.  What will it cost?”  I had two ways of responding to this.  The first way was obnoxious – “Do you want fries with that”.  The second way was more thoughtful – “What are you trying to learn?”

Now that I’m on the vendor side of the equation, I only use the second response.  (The first response isn’t good for business.)  But the question is just as valid as it was 20 years ago – “What are you trying to learn?”  In many cases, it isn’t instantaneously clear what that is and in some cases, it is actually incorrect.  Working with a major consumer packaged goods company, we conducted an A&U trying to understand barriers to consumption for certain buyer groups.  When we got the data together and started to work with the client, it became clear that we weren’t answering the core questions.

It turns out that the core question was more about pricing and perceived price/value than it was about barriers to purchase.  But the client got their marching orders from the brand team and took them at face value and it wasn’t until the data came in and the rubber hit the road that we all realized that we had missed the mark.  Thankfully, the client recognized that the problem wasn’t our fault and we were able to do some additional work to make this meet the client’s strategic needs.  But the experience was a lousy one.

Is there a way to avoid this?  Absolutely!

1. 1.     During the initial conversation about the RFP/potential project, avoid any methodological discussion.
   

•    What sparked the interest in the project?
•    Describe the business issues in as much detail as possible
•    What has changed in the marketplace that has made this an important business issue?
•    What do you know already?
•    What would make this a successful project?
•    Who are the key end users of the results?

1. 2.    Only then should you discuss what the test should look like.
   

•    Method
•    Key Metrics and Decision Criteria
•    Sample

At MarketTools, we are known for our optimization tools.  We constantly challenge ourselves to avoid thinking of these tools as the answers to the world’s questions.  We focus on working better with our clients to make sure we understand what the questions are and what the answers should provide.  If we continue to do this well, the test design is relatively easy to figure out.

1. So, the next time someone asks you if you like his/her clothes, you should ask:

•    What are you really trying to understand ?
•    What are you going to do with the results?

"Social media" is one of those terms that often creates a few immediate reactions: confusion, excitement, and perhaps most often, uncertainty. As companies realize that social media isn’t just the job of the marketing department, many are struggling to understand how this new opportunity to interact with customers fits into their feedback management or voice of customer efforts. 

We see social media as another vital source of feedback within a comprehensive Enterprise Feedback Management (EFM) initiative. Whether you run a global organization with products that touch most households or provide small, niche products for a narrow market, chances are people are talking about your organization – or at least your industry – online. The conversations about your company on Twitter, Facebook, blogs and community forums are rich with insight about what matters most to your target audience.

The challenge with social media feedback is that it’s inherently messy – it’s public, unsolicited and unpredictable. Its “unfiltered” nature can provide more honest conversations – but it can also provoke attention seekers to create real problems for your brand. A comprehensive EFM system should provide the ability to listen to what people are saying on the social web, analyze it to understand the relative importance vs. the noise, and provide a forum for taking action to augment the voice of your promoters and mitigate the potential impact of detractors.

While analyzing customer feedback in social media can yield critical insights, it doesn’t replace proactive customer satisfaction surveys. The private nature of surveys encourages a level of depth and honesty not always offered through social media. And of course, by proactively gathering feedback, you get to ask the questions you want to ask vs. relying on what others want to talk about online.  Survey data is also designed to be easier to slice and dice, trend specific issues over time, and perform advanced analytics such as regressions and significance testing.

The beauty of a solution that allows you to “listen in surround sound” to feedback from surveys and social media is that you can use each type of feedback as a leading indicator. For example, customers may be discussing a topic in a forum that you never thought to ask about in your surveys – allowing you to tune survey questions to ensure you capture feedback about something they consider important. On the flip side, you may uncover an issue within survey results and have the opportunity to fix it before people start discussing it on the social web. To take it even further, you can proactively and publicly communicate how you’ve addressed issues on the social web to let people know your company takes feedback seriously – and stop negative conversations before they get out of control.

So embrace social media as a feedback channel, but don’t lose sight of the value traditional feedback gathering methods can provide as well. The right combination of solicited and unsolicited feedback will give you the best perspective for uncovering what matters most to customers – all the rest is just noise.