Back to Articles|Published on November 19, 2026|30 min read

Tapflare Article

How to Commission an Infographic: Source-to-Claim Sheet

How to commission an infographic in 2026: a 17-field source-to-claim sheet, fictional worked example, blank template, and separate data and rendering sign-offs before design.

By Tapflare

How to Commission an Infographic: Source-to-Claim Sheet

Summary

  1. 01Choose the audience, the decision the reader should make, and one approved message before commissioning design. Use a table when readers need to compare precise values.
  2. 02Give each claim its own row. Record the original publication, exact table or cell, extraction date, and reuse terms before briefing the designer.
  3. 03Check geography, population, period, unit, and denominator before comparing figures. Record the formula, rounding rule, and caveats in the sheet.
  4. 04Brief the designer by claim ID. Have the named data owner approve the figures and a separate reviewer check that the artwork shows them faithfully.
  5. 05Deliver editable files, exports, and text alternatives. Store the updated sheet with the editable file and set a review date.
Inside this article
  1. 01Executive Summary
  2. 02Introduction and Background
  3. 03Choosing the Format: Should This Go to a Designer Now?
  4. 04Auditing Sources: Original Data, Versions and Permitted Use
  5. 05The Source-to-Claim Sheet: Fields, Fictional Example and Blank Template
  6. 06Reconciling and Calculating: Comparability, Percentages and Uncertainty
  7. 07Briefing the Designer and Separating the Two Approvals
  8. 08Delivering Files, Text Alternatives and Updates
  9. 09Data Analysis and Evidence
  10. 10Frequently Asked Questions (FAQs)
  11. 11Implications and Future Directions
  12. 12Conclusion

Executive Summary

Commissioning an infographic is a content decision before it is a design decision. Design Pickle's help center tells clients to supply the infographic text because its designers will not be able to look up or research the content [1], and its scope page lists looking up information about products as unsupported work [2]. Penji's blog likewise says the client gathers and documents the data its designers add [3]. A ManyPixels roundup describes the wider market the same way: per-project services typically handle design only, while higher-end agencies may add data research and copywriting [4]. A buyer therefore needs to know, before ordering, whether the claims are ready for a designer, need an analyst or editor first, or belong in a table or plain text.

This report proposes a source-to-claim sheet: a register with one row per claim and 17 fields (plus a decision column for triage), from claim ID and approved wording through original URL, table or cell, measure, population, period, unit, denominator, extraction date, calculation, caveat, data owner, planned visual encoding and artwork location, ending in two separate sign-offs. It is a specific implementation of existing public guidance, not a claim that nothing similar exists. Statistics Canada's chart guidance says a source should be specific enough for readers to find the data and that the extraction date should be recorded when a source is frequently updated [5]. The UK Office for National Statistics (ONS) asks chart descriptions to state the data used, in a consistent order [6].

The sheet separates two approvals. Source-data approval asks whether the number is traceable and comparable. Faithful-rendering approval asks whether the artwork shows that number without distortion. UK Government Analysis Function guidance warns that incorrect design can produce a misleading infographic [7], and that quality infographics need analysts as well as designers [7]. When a source cannot be traced, or two figures are incomparable, the next action is an evidence decision by the client or editor, not decoration.

A fictional worked example (instructional, not a benchmark) shows the arithmetic discipline: a rate moving from 20% to 30% is +10 percentage points and +50% relative change, calculated as (30 minus 20) divided by 20, times 100. ONS defines a percentage point as the difference between percentages [8]. The report also covers licence and attribution checks, uncertainty, delivered formats with text alternatives under WCAG 2.2 (a W3C Recommendation dated 12 December 2024) [9], and a blank portable sheet. All guidance cited is an attributed documented model, not a universal commercial rule, and a reviewer with statistical training should test any worksheet before relying on it.

17Fields in the source-to-claim sheet before adding the triage column
18Columns in the working register, including the Pass, Hold or Rewrite column
10 percentage pointsPoint increase in scheduling software use in the fictional survey example
+50%Relative increase from the starting rate in the fictional example

Introduction and Background

Buyers of design subscriptions, freelancers and specialist visualization studios often arrive with a spreadsheet, a few screenshots from articles and a deadline. The designer receives them and produces a polished graphic. Whether the numbers in it are right depends on work that happened, or did not happen, before the brief. This report addresses marketing leads at software and e-commerce companies, content teams and agency account managers who order data-heavy infographics, and it frames the decision they face: can this material go to design now, does it need an analyst, researcher or editor first, or would a simpler table or text serve better?

Whether buyers want this exact sheet is untested; nothing here shows that they do.

Existing material covers adjacent ground. Venngage's data-infographic guide says good infographics start with good data and lists free data sources [10]. The ManyPixels roundup discusses supplier models and notes that data visualization requires specific visual skill beyond general graphic design [4]. Statistics Canada's chart guide says a chart should always include the sources of the data used [5], and the ONS service manual requires the source directly after each chart [6]. What a buyer lacks is a handoff artifact linking each visual element to an exact data cell, transformation and named approval.

Disclosure: the publisher of this article, Tapflare, offers a flat-rate monthly subscription for graphic design and website development, managed through a dedicated project manager. It is a design production option, not a research, copywriting or statistical review service, so approved content and data are a dependency it expects the client to supply. That makes it a proportionate fit when the sheet is complete and a poor fit when the claims still need evidence work.

Choosing the Format: Should This Go to a Designer Now?

The first decision is about the message, not the layout. The Rural Health Research Gateway's dissemination toolkit tells creators to identify the take-away and present supporting, accurate data [11], and to use up-to-date reputable sources and cite them [11]. Write the audience, the decision the reader should make and a single approved message in one sentence before opening a design request form.

Then test whether an infographic is the right format. The Analysis Function describes infographics as using charts, timelines, maps or other visualisations to explain data [7], but warns that a simple chart with explanatory text could be clearer and more accessible [7], and that with lots of information to communicate an infographic is probably not the best choice [7]. Its tables guidance says charts suit patterns, trends and relationships [12], whereas a table suits users who want to compare values [12]. A well-structured table can often be as powerful as a chart [13].

Three outcomes follow, and the sheet records which applies to each claim:

  • Send to design now: the claim has a traceable original source, comparable measures, a checked calculation and a named data owner. The brief can reference the sheet row.
  • Needs an analyst, researcher or editor first: a source is a secondary restatement, definitions differ, a denominator is missing or a calculation is unchecked. A designer cannot repair these by drawing. Providers state this boundary themselves, with Design Pickle noting designers will not research content [1].
  • Use a table or text instead: values must be read precisely, the story is a single number, or categories overlap so that a part-to-whole picture would mislead.

A readiness check before hiring covers six items. Hold the approved message and audience, a list of claims, original publication links, usage terms, a named person who can approve the data, and the intended publishing location with its file and accessibility needs. Incomplete data usually surfaces late, as rework, so completing these items first protects the schedule as well as accuracy.

Auditing Sources: Original Data, Versions and Permitted Use

A statistic quoted in an article is a pointer, not evidence. Cornell University Library describes primary sources as the original compilers of data [14], and notes that secondary sources embed data from primary sources to support their own interpretation [14]. A University of Kansas journalism text gives a practical instruction: go to the primary source those media reference [15]. In the sheet, the original publication URL must therefore point to the compiler, such as a statistics office, regulator or survey owner, not to the vendor blog that repeated the figure.

Specificity matters. The Analysis Function says to give the specific data source for each chart and link directly to it where possible [13], and Statistics Canada says a chart should always include the sources of the data used [5]. The US Census Bureau asks users to include the table or file name or number when available [16], and says its citation trail lets readers connect work back to the original underlying data [17]. The sheet's "table/cell or page" field is where that precision lives.

Versions change. Statistics Canada tells chart authors to specify the date data were extracted when the source is frequently updated [5], and its open licence is governed by the terms in force as of the date and time of access [18]. The FORCE11 data citation principles say a citation should allow verifying that the timeslice, version or granular portion retrieved later is the one originally cited [19], and also that whenever a claim relies on data, the data should be cited [19]. DataCite's metadata schema recommends registering a new identifier for a major version change [20]. For an infographic, the extraction date and the version label are small fields that make later updates possible.

Permitted use needs checking per source, because terms differ:

  • UK Open Government Licence v3.0: requires acknowledging the source [21], offers default attribution wording [21] and bars suggesting official status or endorsement [21].
  • Creative Commons Attribution 4.0: bars implying licensor endorsement [22] and requires indicating if the material was modified [23].
  • US Bureau of Labor Statistics: states its publications are in the public domain apart from previously copyrighted photographs and illustrations [24], while asking users to cite it as the source [24].
  • Eurostat: authorises reuse provided the source is acknowledged (Source: ec.europa.eu), and says modifications must be stated clearly to the end user (Source: ec.europa.eu).
  • Statistics Canada Open Licence: supplies "Adapted from Statistics Canada" wording for derived products [18] and says users must not misrepresent the information or its source [18].

The adapted-product wording matters because a designer's chart is a derived product. The Census Bureau tells users who build their own estimates to cite it as the source of the original data only [16], which means the infographic's conclusions belong to whoever approved the calculation, not to the statistics office. This is general information for a commissioning checklist and not legal advice; a licence question that affects publication should go to the client's counsel.

Commissioning an infographic is a content decision before it is a design decision.

The Source-to-Claim Sheet: Fields, Fictional Example and Blank Template

The sheet carries one row per claim and stays attached to the file through approval. Table 1 lists the 17 fields, what to record and why each exists, with the public guidance that motivates it. The measure, geography and period fields follow the ONS chart-description model and are a documented house style, not a mandatory commercial standard.

#FieldWhat to recordWhy it exists
1Claim IDA short unique code (C-01) used in the brief and the artwork layer namesLets a visual element point to a row, not only a source footer
2Approved wordingThe exact sentence, label or number the client approvedPrevents a designer or copy edit from drifting from the verified claim
3Original publication URLThe compiler's page, not a secondary articlePrimary sources are the original compilers of data [14]
4Table/cell or pageTable number, row, column, page or dataset fieldCensus guidance asks for the table or file name or number [16]
5Statistical measureCount, rate, share, mean, indexONS subtitles state the data used [6]
6Population/geographyWho or what is counted, and whereThe Analysis Function subtitle names what the data is, the geography and the period [13]
7PeriodReference year, wave, month rangeSame subtitle guidance; also exposes mismatched windows
8UnitPercent, percentage points, currency, countTable guidance asks that units be described [12]
9DenominatorThe base a percentage is divided byA prevalence figure is a count divided by the total for that same group [25]
10Extraction dateDate the figure was read from the sourceNeeded when the source is frequently updated [5]
11CalculationThe formula and rounding rule, or "none, quoted as published"Keeps transformations visible and checkable
12CaveatDefinition breaks, estimates, suppressed cells, uncertaintyQuality information should not be hidden in footnotes [26]
13Data ownerNamed person who can answer questions about the figureGives the designer and reviewer a human contact
14Planned visual encodingChart type, baseline, labels, colour roleTies appearance to the claim (see rendering review below)
15Artwork locationPanel, layer or frame in the editable fileMakes review and later updates findable
16Source-data sign-offNamed approver, date, statusApproves the number, not the picture
17Faithful-rendering sign-offNamed approver, date, statusApproves the picture, not the number

Table 1 treats source-data approval and rendering approval as separate fields because they answer different questions and are usually held by different people. A single "approved" tick invites the assumption that checking the look also checked the evidence. The sheet adds a decision column (Pass, Hold or Rewrite) on top of the 17 fields for triage, so the working register has 18 columns.

Fictional Completed Example (Hypothetical Example)

Table 2 shows three claims from an invented survey. Every name, number and URL is fictional and instructional: it is not an industry benchmark and not new research. The sheet is a worksheet for a data-literate reviewer to test, and it does not certify statistical validity.

FieldC-01 (Hypothetical Example)C-02 (Hypothetical Example)C-03 (Hypothetical Example)
Approved wording"Use of scheduling software rose from 20% to 30%, up 10 percentage points."Proposed: "Average adoption across both markets is 48.5%.""Top reasons for choosing a tool: price 45%, integrations 38%, support 29%."
Original URLexample.org/sbts/table-4 (fictional)Survey A: example.org/a/t2; Survey B: example.org/b/t7 (fictional)example.org/sbts/table-9 (fictional)
Table/cellTable 4, row "Uses scheduling software", columns 2024 and 2026A: Table 2, cell B3 (42%). B: Table 7, cell D5 (55%)Table 9, rows 1 to 3
MeasurePercentage of respondentsPercentage of respondentsPercentage of respondents selecting each reason
Population/geographySmall businesses with 1 to 49 staff, fictional Country XA: adults in Country X. B: online shoppers in Country YSame as C-01, 2026 wave
PeriodSurvey waves 2024 and 2026A: 2025. B: 20262026 wave
UnitPercentPercentPercent
DenominatorAll respondents in each waveDifferent in each survey (adults versus shoppers)All respondents; multiple answers allowed
Extraction date2026-11-022026-11-022026-11-02
Calculation30 minus 20 = +10 percentage points. Relative: (30 minus 20) / 20 x 100 = +50%Simple average of 42% (A) and 55% (B) = 48.5%, but the inputs are not comparableNone; shares sum to 112 because categories overlap
CaveatQuestionnaire wording must be confirmed identical across wavesA separate Country Z row is "not available" in the source; it must not be drawn as 0Respondents could pick more than one reason
Data ownerClient research lead (named in real use)Client research leadClient research lead
Planned encodingTwo bars, zero baseline, direct labels 20% and 30%, callout "+10 percentage points"Pie or single combined figure (rejected)Originally a pie (rejected)
Artwork locationPanel 2, group "adoption"Not assignedPanel 4, group "reasons"
Source-data sign-offPending, named approver requiredWithheldApproved with caveat
Rendering sign-offNot yet requestedNot requestedPending revised bar chart
DecisionPass to designHold: evidence decision by client or editorRewrite the encoding: horizontal bars with a "more than one answer" note

The decisions in Table 2 show how the sheet works. C-01 passes because both figures share a source, population and denominator, and the calculation is transparent. C-02 holds because two surveys with different denominators cannot be averaged and a missing value is not zero; the client or editor must decide whether to show two separate figures or drop the claim. C-03 has valid source numbers but a flawed encoding plan: overlapping categories cannot be shown as a whole, and the Analysis Function says pie charts need categories that sum to a meaningful whole [13]. A fourth case, not tabulated, is a statistic found only in a vendor blog with no link to an original dataset. The next action is an evidence decision: trace it, replace it with a traceable figure, or drop it. Making the number prettier is not an option.

Blank Portable Sheet

Copy the header row below into a spreadsheet. One row per claim; leave no field blank at sign-off (write "none" or "not applicable" instead).

claim_id,approved_wording,original_url,table_cell_or_page,statistical_measure,population_geography,period,unit,denominator,extraction_date,calculation,caveat,data_owner,planned_visual_encoding,artwork_location,source_data_signoff_name_date,rendering_signoff_name_date,decision_pass_hold_rewrite

Row rules: write approved wording in final form; paste the compiler's URL, not a search result; enter the calculation as a formula, not a result only; mark "estimate" or "unknown" where applicable; and keep the data sign-off and rendering sign-off in different cells with different dates.

Figure 01
Reasons for choosing a tool in the fictional multiple-response surveyPercent of respondents
Source: example.org/sbts/table-9 (fictional)

Reconciling and Calculating: Comparability, Percentages and Uncertainty

Reconciliation happens in the sheet's columns, before design starts. Compare geography, population, period, unit and denominator across every pair of figures that will share a visual.

Comparability. Official guidance is cautious. The Census Bureau advises that comparisons of American Community Survey data across periods be made with caution or not at all when variables or geographies change [27], and to compare 1-year estimates only with 1-year estimates [27]. A European Commission glossary defines a break in series as arising when the standards for defining and observing a variable change (Source: knowledge4policy.ec.europa.eu). ONS has said that reweighted headline labour market statistics are not directly comparable with Labour Force Survey data weighted differently [28]. The sheet's caveat field records such breaks, and the editor decides whether the comparison stands.

Percentages. A percentage is meaningless without its base. Statistics Canada's table guide shows a prevalence percentage as a count divided by the total number of persons in that age group [25]. A change between two percentages is expressed in points, since a percentage point is the difference between percentages [8]. The sheet should store both the point change and, where wanted, the relative change, with the formula visible.

Missing, zero and suppressed values. Statistics Canada distinguishes several symbols: a figure never published for any period [29], a value not logically possible [29], a value suppressed to protect respondent confidentiality [29], and zero, which can be a true zero or a value rounded to zero [29]. Each means something different, and none licenses a designer to draw a bar of zero height.

Rounding. Rounded parts may not match a rounded total. Ofqual's rounding policy notes that the total may be slightly different to the sum of the individual rounded values [30], and Statistics Canada's table guide warns that category sums may differ from the stated total [25]. Put the rounding rule in the calculation field.

Overlapping categories. Multiple-response questions can produce shares above 100. An archived Census 2000 release on race groups shows the pattern, with group sums that exceed 100 percent of the population [31], and Pew Research Center has a policy of using a forced-choice format instead of a select-all-that-apply list in its online surveys whenever possible, because the formats do not yield identical results [32]. Totals are not additive for overlapping categories, and the label must say so.

Uncertainty. Survey figures carry sampling error. Pew describes the margin of sampling error as how close a result can be expected to fall to the true population value [33], warns that the margin for the difference between two polls may be larger than expected [33], and notes that subgroups have larger margins [33]. Error bars show sampling error only and understate total error [34]. ONS advises showing uncertainty when it matters for key trends [35] and leaving it off when it does not affect interpretation [35]. Single-number measures such as standard errors should stay in datasets, not charts [35], and when uncertainty is so high that no meaningful comparison is possible the guidance is not to show the data in a chart [35]. A designer should not invent confidence intervals; the analyst supplies them or the claim is marked unknown.

Briefing the Designer and Separating the Two Approvals

Figure 02
From infographic brief to maintained file
  1. 01Choose the message and format

    Identify the audience, the reader's decision, and the approved message. Test whether an infographic is the right format.

  2. 02Audit the sources

    Trace each figure to its original source and check the version and permitted use.

  3. 03Fill and reconcile the sheet

    Check geography, population, period, unit, and denominator. Document calculations and flag estimates and unknowns.

  4. 04Brief the designer

    Reference every visual element by claim ID and state the relationship it should show.

  5. 05Review the artwork

    Check how the artwork represents the data in a separate pass from data approval and the appearance and copy review.

  6. 06Deliver and set a review date

    Hand over the editable files and text alternatives, and record a date for reviewing updates.

A brief that cites the sheet gives the designer relationships to show, not just numbers to place. State for each row what the viewer should compare: change over time, rank, share of a whole or a single value. Reference every visual element to a claim ID instead of a general source footer. Design Pickle's guide asks for design references, colours and fonts in the request [1], which covers the creative half; the sheet covers the evidence half.

The Process in Seven Steps

The commissioning process, from first idea to maintained file, runs in this order:

  1. Choose the audience, decision and approved message, and test the format.
  2. Audit each source for originality, version and permitted use.
  3. Fill the sheet and reconcile geography, population, period, unit and denominator.
  4. Calculate transformations transparently and mark estimates and unknowns.
  5. Brief the designer by claim ID, with the intended relationship for each element.
  6. Review rendering separately from data and from appearance and copy.
  7. Deliver editable files, text alternatives and an update date.

Who Does What

Responsibilities differ by role, and the sheet makes the handoff explicit:

  • Client or editor: owns the message, approves wording, makes evidence decisions on holds and names the data approver.
  • Analyst or researcher: traces originals, reconciles populations and periods, performs and documents calculations, and marks estimates.
  • Designer: renders approved rows with the planned encoding, keeps layers labelled with claim IDs and flags anything the artwork cannot show faithfully.
  • Reviewer: checks rendering against the rows, separately from the people who approved the data.

Table 3 compares what several suppliers' own pages say about who supplies infographic content, so the buyer can see where analyst work must come from. The pages describe scope as published; terms change, so confirm them at order time. No prices are given here.

Supplier modelWho supplies content/data, per the sourceResearch or copy included?Source
Design PickleClient supplies text; copy can go in the request form [1]Not for research: designers will not look up content [1]; scope lists copy-and-paste text submission [2]Provider help center
PenjiClient gathers and documents the data [3], arranged as it should appear [3]The page assigns data gathering and documentation to the clientProvider blog
SupersideBriefing starts from the client's goals, audience, formats and inspiration [36]The page does not state who sources or verifies data; it describes client review before final production [36]Provider service page
Per-project services and higher-end agenciesClient supplies data and content for per-project services [4]Higher-end agencies may include data research and copywriting [4]ManyPixels roundup (a competitor's description of the market)
TapflareClient supplies approved copy and data through the project manager's request flowDescribed as design and development production; research, copywriting and statistical review are not described as part of the service. Suited to steady design output once the sheet is complete; confirm scope before using it for a one-off job or a strategy engagementPublisher's own description, not independently cited

Table 3 shows a consistent pattern: the common subscription model treats content as a client dependency, and the exceptions sit in higher-end agency tiers whose scope should be confirmed in writing. Tapflare appears in the table as one option on the same factual footing, with its scope described neutrally. If the sheet shows many holds, a supplier that includes analysis or editorial support, or an in-house analyst, is the proportionate next step before any design budget is spent.

Rendering Review, Separate from Appearance and Copy

Review the artwork against the sheet in three passes, each with its own sign-off cell:

  • Data pass (before design): each row's source, version, calculation and comparability are approved by the named data owner.
  • Rendering pass: each visual element matches its row. Bar and area charts start at zero: ONS says an axis for bars should always start at zero [37], Statistics Canada says axes should start at zero and not be broken [5], and the CDC infographics guide says to always begin the y-axis at zero [38]. Readers perceive bar lengths as proportional [13]. If a zero baseline hides the story, one alternative is a dot plot [13]. Avoid dual axes [37], keep comparable charts on the same scale [37], and be cautious with size encodings, since the relative size of dots is hard to estimate [5]. Cropped, stretched or distorted charts give a false impression of the data [39].
  • Appearance and copy pass: typography, brand, tone and wording, checked last and not allowed to change a number without reopening the row.

The unit and source belong on the artwork too: ONS places an axis unit in the axis title [6] and requires the source directly after the chart [6]. NN/g advises that summary statistics show the source used to calculate them [39].

Figure 03
What each infographic approval checks
Source-data approval
  • Record the exact table, row, column, page, or dataset field behind the claim.
  • Compare geography, population, period, unit, and denominator for figures that share a visual.
  • Have the named data owner approve each row's source, version, and calculation before design.
Faithful-rendering approval
  • Check every visual element against its row in the sheet.
  • Check that bar and area charts start at zero.
  • Check that the artwork shows the unit and source.

Delivering Files, Text Alternatives and Updates

Ask for editable source files and an export set. Design Pickle's file-types page lists the finished formats it delivers [40]; check the equivalent for any supplier. Ask for the sheet to be returned updated with artwork locations.

Accessibility is a delivery requirement, not an afterthought. WCAG 2.2 Success Criterion 1.1.1 requires a text alternative serving the equivalent purpose [9], and the W3C complex-images tutorial says long descriptions should cover scales, values, relationships and trends [41] and may describe structure such as the sequence of colours and relative column heights [41]. It recommends making long descriptions available to everyone [41] and referring to and summarising complex images from the surrounding text [41]. Where the long description contains a table, the tutorial notes that its aria-describedby approach only works for text-only long descriptions [41]. Colour must not be the only means of conveying information [42], so series encoded by colour need labels. This report does not reproduce the WCAG matrix or choose palettes; consult the current W3C documents and the publishing platform's help pages.

Three output options suit different needs:

  • Image plus body-text description and data table: the W3C decision tree says to include the information contained in the image elsewhere on the page [43], and a technique for line graphs places the data table used to generate the graph in the long description [44].
  • Table-first, chart second: for readers who need values, the Analysis Function says to give a table if users are expected to read data from the chart [13]; the text alternative may be a table, a description or both [45].
  • Native live chart: the claim wording and source still belong in page text, since Analysis Function guidance warns that titles, source information and footnotes placed inside an image may fail accessibility requirements [45].

Do not publish a table as a flattened picture [12], and avoid embedding information only inside an image [46]. Platform documentation matters in practice: Microsoft's guidance asks alt text to describe the object and its context [47], and WordPress documentation says to describe the image or leave alt text empty if decorative [48], noting that library alt text sets the default for later inserts [48]. Provide data downloads where possible, since they improve transparency [45].

For maintenance, store the sheet, the extraction dates and the editable file together, and note a review date. When a source updates, the data owner decides whether each affected row is refreshed, held or retired.

A single "approved" tick invites the assumption that checking the look also checked the evidence.

Data Analysis and Evidence

This report draws on published guidance and not on original measurement, so its evidence is about what documented sources say, plus transparent arithmetic. No public benchmark of error rates in commissioned infographics was found, and none is claimed.

Worked calculations (all fictional and instructional, not benchmarks).

  • Point versus relative change: a rate moving from 20% to 30% is +10 percentage points. Relative change is (30 minus 20) divided by 20, times 100, which is +50%. It is not a 10% rise, and a label must say which is meant.
  • Averaging incomparable surveys: (42 + 55) / 2 = 48.5 looks arithmetic but combines two denominators, so the result describes no population.
  • Overlapping categories: 45 + 38 + 29 = 112, which cannot be shown as slices of one whole.
  • Missing versus zero: a "not available" cell is unknown; entering 0 changes any sum and average.
  • Rounding: three shares of 33.3% sum to 99.9%, and the sheet's rounding rule decides whether a total is shown as 100.

What the sources establish. Of the supplier pages read, Design Pickle, Penji and the ManyPixels roundup all place data supply with the client, with the ManyPixels page allowing that higher-end agencies may include research. That is three published statements, from three providers, not a survey of the market. ONS and the Analysis Function describe documented house styles for government communication; they are useful models and not mandatory commercial standards. Statistics Canada's guide carries a February 24, 2023 release date [5], and the WCAG 2.2 Recommendation December 12, 2024 [9], so each should be checked for later revisions before reuse. The Design Pickle page read for this report displays a September 28, 2026 date, which may differ from other dates seen for the guide.

Demand. No public data cited in this report shows that buyers want this exact sheet.

Limits. The sheet is untested with real buyers. Sample sources may change terms. A reviewer with statistical training should check the worksheet and its arithmetic, since a reviewed worksheet shows the process is traceable, not that the underlying analysis is valid.

Frequently Asked Questions (FAQs)

How do you commission an infographic? Settle the message and audience, test whether a table or text would serve, complete the sheet for every claim, resolve holds, then brief the designer with claim IDs. Seven steps: choose message, audit sources, fill the sheet, reconcile and calculate, brief, review rendering separately, deliver with text alternatives and an update date.

What goes in an infographic design brief template? Message, audience, format and placement, brand references, and the source-to-claim sheet for evidence. The sheet above is the data half of the brief.

How do you brief a designer for an infographic? Name the relationship each row should show, point each visual element to a claim ID and state sign-off owners. Provider request guides may also ask for design references and brand details.

What should a buyer prepare before hiring an infographic designer? Approved wording, original links, usage terms, a data owner and a decision on format. The Analysis Function notes quality infographics need analysts and designers together [7].

What are good data sources for infographics? Originators: statistics offices, regulators and survey owners, such as the Census Bureau, BLS, Eurostat and Statistics Canada, each with its own reuse terms. Venngage notes that Census tables can be downloaded in Excel or other formats [10]. Check licence and citation wording per source.

How should infographic sources and citations appear? Name the specific publication and link to it [13], with the extraction date where the source updates [5]. Use licence wording such as the OGL default statement [21] where required.

How do you fact check an infographic? Separate the checks. Verify each claim against the original, recompute each calculation, confirm comparability and then check that the artwork shows the number without distortion. Digital.gov suggests verifying data are complete, consistent and standardized [49] and identifying limits in how the data were collected [49].

What belongs on an infographic content checklist? A readiness checklist: wording, source URL, cell, measure, population, period, unit, denominator, extraction date, calculation, caveat, owner, encoding, artwork location and two sign-offs.

Implications and Future Directions

For buyers, the practical implication is that the sheet moves a quality conversation upstream, to a point where changes are cheap. Editing a number in a spreadsheet is cheaper than redrawing a finished panel, and a supplier that cannot take the sheet as input is signalling that its scope excludes evidence work, which is useful information before purchase.

For suppliers, a standard handoff reduces ambiguity about responsibility. Public guidance points toward shared accountability: the Analysis Function says high-quality infographics need a team including analysts and designers [7], and the Census Bureau frames citation as enabling discovery, reuse and replication [50]. A supplier can receive the sheet as a precise statement of what has and has not been verified.

Several open questions remain. The sheet is long for small projects, and a reduced version covering source, measure, period, denominator and sign-off may be proportionate for a single-number graphic. Automation could populate fields from dataset metadata, but data owners would still need to approve the extraction. Tooling and platform accessibility support also change, so text alternatives should be rechecked against current platform documentation. No evidence yet shows that educators or design-brief guides will adopt such a worksheet; that is a hypothesis.

Conclusion

An infographic order is ready for design when each claim has a traceable original source, comparable units and populations, a checked calculation and a named data approver. It needs an analyst, researcher or editor first when sources are secondary, definitions differ, denominators are missing or uncertainty is unexplained. It should become a table or text when readers must compare precise values, when categories overlap or when a simple chart with explanation is clearer.

The source-to-claim sheet makes those judgments visible. Its 17 fields (plus a decision column for triage) connect approved wording to the exact table, measure, population, period, unit, denominator, extraction date and calculation behind it, and it keeps source-data approval apart from faithful-rendering approval. Hold, rewrite and pass decisions give the client or editor the next action when evidence is weak, which is an evidence decision and not a styling decision.

Supplier scope varies. The pages cited here place content and data with the client in common subscription models, so a buyer should confirm in writing who does research, copy and statistical review. Where the sheet is complete, a subscription design service (including Tapflare, the publisher of this article) can be a proportionate production option; where it is not, evidence work comes first. Test the sheet with a data-literate reviewer, adapt the fields to the project, and keep it with the editable files for the next update.

External Sources (50)

About

Tapflare

Tapflare in a nutshell Tapflare is a subscription-based “scale-as-a-service” platform that hands companies an on-demand creative and web team for a flat monthly fee that starts at $649. Instead of juggling freelancers or hiring in-house staff, subscribers are paired with a dedicated Tapflare project manager (PM) who orchestrates a bench of senior-level graphic designers and front-end developers on the client’s behalf. The result is agency-grade output with same-day turnaround on most tasks, delivered through a single, streamlined portal.

How the service works

  1. Submit a request. Clients describe the task—anything from a logo refresh to a full site rebuild—directly inside Tapflare’s web portal. Built-in AI assists with creative briefs to speed up kickoff.
  2. PM triage. The dedicated PM assigns a specialist (e.g., a motion-graphics designer or React developer) who’s already vetted for senior-level expertise.
  3. Production. Designer or developer logs up to two or four hours of focused work per business day, depending on the plan level, often shipping same-day drafts.
  4. Internal QA. The PM reviews the deliverable for quality and brand consistency before the client ever sees it.
  5. Delivery & iteration. Finished assets (including source files and dev hand-off packages) arrive via the portal. Unlimited revisions are included—projects queue one at a time, so edits never eat into another ticket’s time.

What Tapflare can create

  • Graphic design: brand identities, presentation decks, social media and ad creatives, infographics, packaging, custom illustration, motion graphics, and more.
  • Web & app front-end: converting Figma mock-ups to no-code builders, HTML/CSS, or fully custom code; landing pages and marketing sites; plugin and low-code integrations.
  • AI-accelerated assets (Premium tier): self-serve brand-trained image generation, copywriting via advanced LLMs, and developer tools like Cursor Pro for faster commits.

The Tapflare portal Beyond ticket submission, the portal lets teams:

  • Manage multiple brands under one login, ideal for agencies or holding companies.
  • Chat in-thread with the PM or approve work from email notifications.
  • Add unlimited collaborators at no extra cost.

A live status dashboard and 24/7 client support keep stakeholders in the loop, while a 15-day money-back guarantee removes onboarding risk.

Pricing & plan ladder

PlanMonthly rateDaily hands-on timeInclusions
Lite$6492 hrs designFull graphic-design catalog
Pro$8992 hrs design + devAdds web development capacity
Premium$1,4994 hrs design + devDoubles output and unlocks Tapflare AI suite

All tiers include:

  • Senior-level specialists under one roof
  • Dedicated PM & unlimited revisions
  • Same-day or next-day average turnaround (0–2 days on Premium)
  • Unlimited brand workspaces and users
  • 24/7 support and cancel-any-time policy with a 15-day full-refund window.

What sets Tapflare apart

Fully managed, not self-serve. Many flat-rate design subscriptions expect the customer to coordinate with designers directly. Tapflare inserts a seasoned PM layer so clients spend minutes, not hours, shepherding projects.

Specialists over generalists. Fewer than 0.1 % of applicants make Tapflare’s roster; most pros boast a decade of niche experience in UI/UX, animation, branding, or front-end frameworks.

Transparent output. Instead of vague “one request at a time,” hours are concrete: 2 or 4 per business day, making capacity predictable and scalable by simply adding subscriptions.

Ethical outsourcing. Designers, developers, and PMs are full-time employees paid fair wages, yielding <1 % staff turnover and consistent quality over time.

AI-enhanced efficiency. Tapflare Premium layers proprietary AI on top of human talent—brand-specific image & copy generation plus dev acceleration tools—without replacing the senior designers behind each deliverable.

Ideal use cases

  • SaaS & tech startups launching or iterating on product sites and dashboards.
  • Agencies needing white-label overflow capacity without new headcount.
  • E-commerce brands looking for fresh ad creative and conversion-focused landing pages.
  • Marketing teams that want motion graphics, presentations, and social content at scale. Tapflare already supports 150 + growth-minded companies including Proqio, Cirra AI, VBO Tickets, and Houseblend, each citing significant speed-to-launch and cost-savings wins.

The bottom line Tapflare marries the reliability of an in-house creative department with the elasticity of SaaS pricing. For a predictable monthly fee, subscribers tap into senior specialists, project-managed workflows, and generative-AI accelerants that together produce agency-quality design and front-end code in hours—not weeks—without hidden costs or long-term contracts. Whether you need a single brand reboot or ongoing multi-channel creative, Tapflare’s flat-rate model keeps budgets flat while letting creative ambitions flare.

Disclaimer

This document is provided for informational purposes only. No representations or warranties are made regarding the accuracy, completeness, or reliability of its contents. Any use of this information is at your own risk. Tapflare shall not be liable for any damages arising from the use of this document. This content was generated with assistance from artificial intelligence tools, which may contain errors or inaccuracies. Readers should verify critical information independently. All product names, trademarks, and registered trademarks mentioned are property of their respective owners and are used for identification purposes only. Use of these names does not imply endorsement. This document does not constitute professional or legal advice. For specific guidance related to your needs, please consult qualified professionals.