Use Cases: Building Surveys with Templates, Logic & AI

Combining Templates, Logic, & AI

Crafting insightful surveys by weaving together templates, logic, and AI can be quite a challenge. Let's break it down . . .

Great survey questionnaires — from the high-stakes to the fanciful — are anchored by great questions, with metrics designed to reduce bias.

Some strategic question types have perplexing names — like net promoter (NPS), customer satisfaction (CSAT), customer effort (CES), and the iconic Likert scale — pictured here. Each of these proven question types have been built into Surveys' templates.

Other methods are seemingly lighthearted — like thumbs, smileys, and 5 stars. These playful names can be misleading; they are legitimate metric-driven strategies with their own valued place in the survey space.

Start with a Surveys Template!

Delighted Surveys offers a wide selection of curated, pre-designed templates to jump start your research. Each template adheres to the 5 keys to effective questionnaires presented in the next section.

Learn the details in Creating Surveys from Templates — Best Practice.

Enhance your Templates with Logic and AI inspired questions

In a well-build questionnaire, insightful follow-up questions are required. For instance, a carefully worded text question can be particularly helpful if tactically positioned after a conditional logic question.

Segment your respondents with logic and a bit of AI

With logic, positive respondents (4-5 stars) can answer a different set of follow-up questions than their less enthusiastic counterparts, who click (1-3 stars).

Creating great follow-up questions shouldn't be a headache. Use AI Recommended Questions to pinpoint great follow-ups for each of your segmented groups of respondents (ex. 1-3 stars or 4-5 stars).

5 keys to effective questionnaires

And with that as our preview, let’s jump into five essential guidelines leading to effective and impartial questionnaires.

Apply purposeful follow-up questions — with a touch of AI
Consider the applicable survey types and metrics
Add logic to segment unique respondents
Visualize your reporting — right from the beginning
Determine the sample size for your questionnaire

1. Apply purposeful follow-up questions — with a touch of AI

Delighted Surveys lets you create every possible question type you can envision. Pinpointing the purpose of each question is the first order of business.

Start by asking:

What do I wish to measure?

While this question is a bit of a cliche in survey circles, this fact doesn’t diminish the importance of clearly defining what it is that you wish to measure.

Is it:

Attitudes, opinions, or beliefs?
Satisfaction?
Loyalty?
Effort or friction?
Employee engagement?
Demographics?
Reactions to the catered BBQ at the annual museum fundraiser?

Pinpointing your purpose makes writing great follow-up questions much easier.

Avoid 7 mistakes

For a refresher on "tight" question writing, visit Avoiding biased questions: 7 examples of bad survey questions from our Delighted Blog.

To summarize this post:

Avoid leading questions
Avoid loaded questions
Avoid double-barreled questions
Avoid jargon
Avoid double-negatives
Sync answer scales with your question's metrics — as explained next in part 2. Consider the applicable survey type and metric
Format your questionnaire for every device collecting feedback. (Luckily, Delighted does this bit for you!)

Avoid 7 mistakes with templates and a touch of AI

Surveys' templates have all been vetted to avoid the 7 mistakes. They'll get you off to a proper start. But remember, you can always swap out questions in templates if you find a better one — if you do so before you publish.

On that note, Delighted's AI Recommended Questions can help refine your follow-up questions — offering you three, on-point alternatives at every go.

The only rub is that AI requires 2 questions to establish the target audience and your survey's purpose. Fortunately, nearly every Surveys' template already has 2 or more questions, so you can apply AI Recommended alternatives right away. Learn the details in Applying AI Recommended Questions — Best Practice.

2. Consider the applicable survey type and metric

Identifying a clear purpose makes selecting appropriate question metrics a snap.

Over the past century, specific methods have emerged that help researchers present focused questions that are less prone to unintentional bias and that can fit almost any research scenario. You'll see these question types built into Surveys' templates.

Question types & metrics

Purpose	Alternatives
Attitudes, opinions, or beliefs	5-point Likert scale or graphical 5 Star question types are good fits
Satisfaction	5-point Customer Satisfaction (CSAT) with labels from 1: Very dissatisfied to 5: Very satisfied will perform well, as can a graphical Smileys question that drives a measurable CSAT response
Loyalty	A Net Promoter Score® or NPS® question with an 11 point loyalty metric segmenting respondents into — (0-6) Detractors, (7-8) Passives, and (9-10) Promoters is the likely choice
Effort or friction	5 or 7-point Customer Effort Score (CES) question with labels from1: Strongly disagree to 5/7: Strongly disagree can hit the mark
Employee engagement	Employee Net Promoter Score (eNPS) applies a 0-10 point NPS scale to internal employee metrics
Demographics	Try our Recommended questions in Surveys for a dozen ideas
Reactions to the catered BBQ at the annual museum fundraiser	A lighthearted and carefully worded 5 star or Smileys survey will work providing approachable versions of Likert scale, CSAT, or CES metric types

(Visit 10 prominent survey methods and proven question types for the deep-dive details and examples of each.

3. Add logic to segment unique respondents

Logic lets you create questions that branch respondents to unique follow-up questions based on their selections.

For example, with a 5 star satisfaction question, you can:

Direct unhappy respondents (1-3 stars) to a follow-up text question
Allow mostly satisfied respondents (4 stars) to land on a Thank you
Send super satisfied respondents (5 stars) to a testimonial request

Hovering over logic icons — so you don't get lost

A logic icon tags the upper corner of every branched question. The icon will help you remember that it's tied to a logic-triggering question above it in the survey flow. Hover over the icon to read the conditional question. Clicking on this icon will take you directly to the logic-triggering question so you can inspect it.

Learn all about applying logic

Visit Adding Logic to Questions to all about this essential filtering feature.

Back

4. Visualizing the Reporting right from the beginning

To help you anticipate, take a peek at the end of the process; check out how the results can be visualized in Delighted Surveys:

1. Average (mean)— calculations and visualizations

2. Top 2 Box — calculations and visualizations

3. NPS & eNPS — calculations and visualizations

4. Yes/No (vote) — calculations and visualizations

5. Multiple choice — visualizations

6. Text visualizations — visualizations

5. Determining a sample size for your survey questionnaire

At Delighted, we like to make things simple — so we created a Sample Size calculator that you can use to estimate the number of people you’ll need to survey to collect accurate results. ("Sample size" is the number of people you should survey to gather results within a margin of error.)

Click here to try the Delighted Sample Size calculator!

Formulas are applied to determine the appropriate sample size for a survey. A common formula is:

Breaking down the formula

There are a number of factors to consider, including:

The population — What is the size of the overall population the you are surveying? This might be your customer base, or the number of employees at your company.
The margin of error — What is the margin of error (ex. 5%) that you are willing to accept?
The confidence level — What is the confidence level you need (ex. 95%) to verify that your results are accurate?

n = (Z^2 * p * (1-p)) / E^2

Where:

n = sample size
Z = z-score for the desired confidence level (1.96 for a 95% confidence level—the calculator will grab this number for you!)
p = population proportion (estimate of the percentage of respondents with a particular characteristic)
E = margin of error (5% in this case)

Glance at the example entered into the image of Delighted's sample size calculator above. In this, we imagined that 320 people attended a Museum Fundraiser. We want to know how satisfied they were with the event. And we wanted a 95% confidence level with a margin of error of 5%.

Plugging that information into the formula, the sample size would be 175.

Now, with all that sorted, its time to polish your questionnaire!

10 prominent survey methods and proven question types

	Low scale	High scale	Description
1. Likert and 5 stars	Strongly disagree, Very unhappy, Not important, Very poor, 1 star	Strongly agree, Very happy, Very Important, Very good, 5 stars	The Likert scale (normally 5 or 7-points) is used to measure the intensity of opinions, attitudes, beliefs, perceptions, or feelings. Named after its inventor, psychologist Rensis Likert, the scale appeared in 1932, making it the OG of survey methods. (Graphical 5 Stars may be used as lighter weight substitute to the Numeric method.) See the calculation and a visualizations in 1. Average (mean).
2. NPS	0 - Not likely	10 - Very likely	Net Promoter Score®, or NPS®, delivers a loyalty metric using an 11 point scale, from 0-10, that segments respondents into three groups — Detractors, Passives, and Promoters.Created in 2003 by Fred Reichheld at Bain & Company, this methodology has been shared so anyone can apply it. Bain also generates benchmarks for companies to measure against their NPS. See the calculation and visualizations in 3. NPS & eNPS.
3. CSAT and Smileys	1 - Very dissatisfied, 1 smile	5 - Very satisfied, 5 smiles	Customer Satisfaction or CSAT is usually a 5-point scale — with labels from 1: Very dissatisfied to 5: Very satisfied.Evolving to its current form in the 1970s, CSAT is so well understood that labels are less necessary than in the past. This means that Smileys can work well if the question drives a measurable CSAT response. Smileys trend toward increased response rates and slightly higher scores. See the calculations and visualizations in 1. Average (mean) and 2. Top 2 Box.
4. CES and Smileys	1 - Strongly disagree	5 - Strongly agree	Customer Effort Score or CES are usually driven by 5-point or 7-point scales — where labels 5 or 7 = Strongly agree and 1 = Strongly disagree.Developed in 2010 by the Corporate Executive Board (now Gartner), CES measures “effort” as a key driver of repeat customers, loyalty, and word of mouth. A Smileys method works well if the question text drives a measurable CES response. See the calculations and visualizations in 1. Average (mean) and 2. Top 2 Box.
5. eNPS	0 - Not likely	10 - Very likely	Employee Net Promoter Score or eNPS uses a 0-10 point scale.Created to help measure employee loyalty and engagement within companies and organizations, eNPS uses the NPS scoring system which segments employees into three groups — Detractors, Passives, and Promoters. See the calculation and visualizations in 3. NPS & eNPS)
6. PMF	1 - Not disappointed	3 - Very disappointed	Product Market Fit or PMF asks, “How would you feel if you could no longer use [brand/product name]?” with a 3-point emotion scale — ex. 1: Not disappointed, 2: Mildly disappointed, 3: Very disappointed.PMF is most often presented in a 3-point Smileys format, and should appear with labels to avoid confusion. See the calculations and visualizations in 1. Average (mean) and 2. Top 2 Box)
7. Text	Open	Open	Text types include follow-up questions, “Other” list/text options, and open comments. Open responses are considered essential research tools that can help pinpoint the roots and depth of a response. See 6. Text visualizations.
8. Multiple choice	Open	Open	Select one or more options — ex. Multiple choice lists, Multiple Selection lists, None of the above, All of the above, True-False, True-Somewhat-False, Yes-No, “Other” text option comments, Free/open text comments, and more. See 5. Multiple choice visualizations.
9. Graphic — Thumbs	Down	Up	Thumbs surveys offer all the simplicity of a clear-cut vote — which is handy and unambiguous in many circumstances. However, with no scale to check for nuance, or with the lack of a follow-up text question to dig deeper into the why, the results are prone to the immediate biases of the moment. See the calculations and visualizations 4. Yes/no or Vote — Thumbs.
10. Graphic — 5 stars and Smileys	1 star, Angry smiley	5 stars, Happy smiley	The highly popular 5 star and Smileys survey methods work as flexible, lightweight, and approachable versions of a Likert scale, CSAT, CES, or other survey options.5 stars and Smileys trend toward increased response rates and slightly higher scores than their numeric counterparts. See the calculations and visualizations in 1. Average (mean) and 2. Top 2 Box)

View calculations and visualizations

1. Average (mean)

Mean=(Sum of all scores)/(Number of responses)

Scores are rounded to the nearest tenth

Visualization samples include:

5 Stars as an average — applies also to Likert scale
5-point CSAT as an average
7-point CES as an average — applies also to Likert scale
3-point PMF as an average

5 Stars as an average

5 Stars can be viewed as either a bar chart or as a five-star graphic — which calculates and displays an average.

Score total: 3(5) + 2(4) + 1(3) + 0(2) + 1(1) = 27

Response total: 3 + 2 + 1 + 0 + 1 = 7

27 / 7 = 3.9

5-point CSAT as an average

Score total: 5(5) + 4(4) + 2(3) + 1(2) + 2(1) = 51

Response total: 5 + 4 + 2 + 1 + 2 = 14

51 / 14 = 3.642 which rounds to 3.6

7-point CES as an average

Score total: 0(7) + 1(6) + 1(5) + 2(4) + 1(3) + 1(2) + 3(1) = 27

Response total: 0 + 1 + 1 + 2 + 1 + 1 + 3 = 9

27 / 9 = 3

3-point PMF as an average

Score total: 4(3) + 1(2) + 2(1) = 16

Response total: 4 + 1 + 2 = 7

16 / 7 = 2.3

2. Top 2 Box

(Total number of “satisfied/happy” responses / Total number of responses) x 100 = CSAT, CES, or Smileys 'happiness' score

Visualization samples include:

5-point CSAT Top 2 Box for a “Satisfaction” score
5-point Smileys Top 2 box for a “Happiness” score
3-point Smileys Top 1 Box
PMF Top 2 Box

5-point CSAT Top 2 Box for a “Satisfaction” score

(9/14)*100 = 64

5-point Smileys Top 2 box for a “Happiness” score

(5/7)*100 = 71

3-point Smileys Top 1 Box

(3/6)*100 = 50

PMF Top 2 Box

(5/7)*100 = 71

3. NPS & eNPS

NPS surveys begin with a carefully crafted question:

“How likely are you to recommend [your company/product] to a friend?”

Respondents provide their answers on a scale from 0 to 10, which segments them into three groups:

Promoters: 9-10
Passives: 7-8
Detractors: 0-6

The net promoter equation for both the NPS and eNPS methods is:

% Promoters - % Detractors = NPS (or eNPS)

NPS rounds to a whole number

Visit our Delighted NPS calculator to better understand this unique scoring system.

4. Yes/no or Vote — Thumbs

Thumbs can be visualized with thumbs or with ring chart.

% Yes (or Up) / % No (or Down)

5. Multiple choice

Visualization sample:

6. Text