{"id":3599,"date":"2026-08-24T00:00:00","date_gmt":"2026-08-23T22:00:00","guid":{"rendered":"https:\/\/humes.pl\/delphi-method-how-to-develop-a-forecast-based-on-expert-consensus-when-hard-data-are-unavailable\/"},"modified":"2026-08-25T15:34:39","modified_gmt":"2026-08-25T13:34:39","slug":"delphi-method-how-to-develop-a-forecast-based-on-expert-consensus-when-hard-data-are-unavailable","status":"publish","type":"post","link":"https:\/\/humes.pl\/en\/delphi-method-how-to-develop-a-forecast-based-on-expert-consensus-when-hard-data-are-unavailable\/","title":{"rendered":"Delphi method: how to develop a forecast based on expert consensus when hard data are unavailable"},"content":{"rendered":"<p>You need to decide whether to enter a new segment, assess the pace of technology adoption, or estimate how a market will develop when it has no sales history or time series to model. Traditional quantitative forecasting has limited usefulness where there is no data to extrapolate, and it is precisely in such situations that <strong>Delphi method forecasting<\/strong> becomes a tool that turns dispersed expert knowledge into a structured, repeatable estimate. This is not guesswork &#8211; it is a designed process for arriving at a well-founded consensus or a stable distribution of assessments.<\/p>\n<h2>When does Delphi method forecasting work better than a quantitative model?<\/h2>\n<p>The Delphi method (a Delphi study) was developed as a qualitative forecasting technique for areas where hard historical data is lacking but dispersed specialist knowledge exists. Rather than relying on time series, it uses structured communication among experts across several rounds to gradually bring their assessments closer to a shared, well-founded position or to better describe the range of differences.<\/p>\n<p>Forecasting using this method makes sense primarily in several typical situations. Below are the most common contexts in which Hume&#8217;s Institute recommends a Delphi study instead of quantitative modeling:<\/p>\n<ul>\n<li><strong>Markets without a history<\/strong> &#8211; new product categories, emerging technologies, and services that are only beginning to take shape and have no sales data for trend analysis.<\/li>\n<li><strong>Long-term horizon<\/strong> &#8211; forecasts extending several or more than a dozen years ahead, where extrapolating the past is unreliable due to expected structural changes.<\/li>\n<li><strong>Phenomena that are difficult to quantify<\/strong> &#8211; regulatory changes, technological breakthroughs, and shifts in social behavior that are difficult to reliably capture in measurable variables.<\/li>\n<li><strong>High uncertainty and conflicting signals<\/strong> &#8211; situations in which the available partial data leads to divergent conclusions and expert interpretation is needed.<\/li>\n<\/ul>\n<p>A key feature of the Delphi method is that forecasts are built on qualitative judgment but structured quantitatively &#8211; through the aggregation of assessments, measures of dispersion, and controlled convergence of positions. This distinguishes it from a standard panel discussion, which has no mechanism for controlling the convergence of assessments.<\/p>\n<h2>How does a Delphi study proceed step by step?<\/h2>\n<p>Delphi forecasting is based on several rounds of surveys conducted with the same panel of experts, with anonymized responses and feedback between rounds. Below is the process structure that Hume&#8217;s Institute uses in forecasting projects based on expert consensus:<\/p>\n<ol>\n<li><strong>Defining the problem and forecasting questions.<\/strong> The research team formulates precise questions &#8211; preferably ones that experts can answer with a numerical value or an unambiguous assessment, such as the year an adoption threshold will be reached, a range of market share, or the probability of an event.<\/li>\n<li><strong>Selecting the expert panel.<\/strong> This decision has the greatest impact on forecast quality. The expert panel should combine different perspectives: market practitioners, researchers, and representatives of different parts of the value chain. The aim is diversity of knowledge sources, not unanimity.<\/li>\n<li><strong>First round.<\/strong> Experts respond individually and anonymously, often providing justification. Both numerical assessments and the arguments behind them are collected.<\/li>\n<li><strong>Analysis and feedback.<\/strong> The team calculates central tendency and dispersion measures for the responses, then shares them with participants along with anonymous arguments &#8211; especially those that depart from the mainstream view.<\/li>\n<li><strong>Subsequent rounds.<\/strong> Experts revise their assessments in light of the arguments of other participants. Those who retain extreme responses are asked to provide additional justification. Rounds are repeated until the distribution of responses stabilizes.<\/li>\n<li><strong>Closure and report.<\/strong> Expert consensus does not mean complete agreement, but rather convergence of assessments within an acceptable range, alongside an understanding of the sources of any differences.<\/li>\n<\/ol>\n<p>The core of the method is anonymity and controlled feedback. They enable experts to change their minds in response to an argument rather than the authority or position of another person in the room. As Hume&#8217;s Institute experts point out, an individual specialist is often wrong, and in an open discussion involving ten people at once, it is usually not the best argument that prevails but the loudest voice &#8211; Delphi organizes this noise into a credible consensus by separating the strength of an argument from the strength of personality.<\/p>\n<p>In Hume&#8217;s Institute projects, the greatest value is observed not in the first round but between rounds &#8211; when an expert sees the rationale behind an assessment that differs from their own and has to respond to it. This mechanism is what distinguishes qualitative forecasting using the Delphi method from simply averaging opinions.<\/p>\n<h2>What mistakes should be avoided, and how does Delphi differ from other methods?<\/h2>\n<p>The Delphi method is methodologically robust, but it is sensitive to several common design errors. Below are the pitfalls that most often reduce the quality of forecasts based on expert consensus:<\/p>\n<ul>\n<li><strong>Poorly selected panel.<\/strong> If experts represent one school of thought or one part of the market, consensus will emerge quickly but will be affected by a shared systematic bias. Panel diversity is a safeguard, not an obstacle.<\/li>\n<li><strong>Imprecise questions.<\/strong> The question &#8220;Will the market grow?&#8221; is not suitable for Delphi. The question must require an answer that is comparable across experts and rounds.<\/li>\n<li><strong>Forcing artificial agreement.<\/strong> The goal is not to bring everyone to a single number. Persistent divergence can be valuable information about real uncertainty &#8211; concealing it distorts the picture.<\/li>\n<li><strong>Panel fatigue.<\/strong> Too many rounds lead to declining engagement and study dropout. The process should be closed when assessments stabilize, rather than prolonged indefinitely.<\/li>\n<li><strong>Lack of high-quality feedback.<\/strong> If only numbers are shared between rounds without arguments, the method degenerates into anonymous voting and loses its learning mechanism.<\/li>\n<\/ul>\n<p>It is also worth understanding how the Delphi method differs from related approaches. A standard expert panel or group workshop provides quick conclusions, but is susceptible to the dominance of the strongest personalities and groupthink. Traditional quantitative forecasting is most useful where data is available, but has limited usefulness for phenomena with no history. The Delphi method sits between them: it is qualitative forecasting with procedural rigor.<\/p>\n<p>In research practice, Delphi is sometimes combined with other techniques &#8211; for example, the results of a Delphi study may inform scenarios or serve as input assumptions for a quantitative model. The method does not replace data where it exists; it complements it where it is lacking. Hume&#8217;s Institute treats it as part of a broader forecasting toolkit, rather than a universal solution to every question about the future.<\/p>\n<h2>How can you assess whether a Delphi study is right for your question?<\/h2>\n<p>Before deciding to launch a study, it is worth checking whether your forecasting question is suitable for this method. Below is a checklist to help assess how well the Delphi method fits the problem:<\/p>\n<ul>\n<li><strong>Is historical data lacking?<\/strong> If you have reliable time series, consider quantitative modeling first.<\/li>\n<li><strong>Is there a distinct group of experts?<\/strong> The method requires access to specialists with real knowledge of the phenomenon, not random respondents.<\/li>\n<li><strong>Can the question be formulated in measurable terms?<\/strong> Forecasting works best for questions about values, ranges, timeframes, or probabilities.<\/li>\n<li><strong>Is there time for several rounds?<\/strong> A Delphi study is a process spread over time, not a single survey.<\/li>\n<li><strong>Do you accept a result in the form of a range and distribution?<\/strong> The method provides a well-founded consensus or a stable distribution of assessments with a measure of uncertainty, rather than one &#8220;certain&#8221; number.<\/li>\n<\/ul>\n<p>If the answers indicate a good fit, the Delphi method will allow you to turn dispersed expert knowledge into a forecast that can be documented, defended, and updated in a subsequent cycle.<\/p>\n<h2>Frequently asked questions<\/h2>\n<h3>What is the Delphi method?<\/h3>\n<p>The Delphi method is a qualitative forecasting technique in which a group of experts anonymously answers the same questions over several rounds. After each round, participants receive aggregated results and the arguments of others, then revise their assessments. The process is repeated until responses stabilize, providing a well-founded picture of consensus or divergence along with a measure of dispersion.<\/p>\n<h3>How does Delphi differ from a standard expert panel?<\/h3>\n<p>A standard panel is based on open discussion, where the strongest personalities can easily dominate and groupthink can emerge. Delphi introduces anonymity and controlled feedback between rounds, meaning that the strength of the argument matters rather than the participant&#8217;s position or volume. This makes consensus more resistant to social biases.<\/p>\n<h3>How many rounds are needed to reach consensus?<\/h3>\n<p>In practice, two to three rounds are usually sufficient, as assessments tend to stabilize relatively quickly. However, the number of rounds is not predetermined &#8211; the process is closed when the distribution of responses stops changing significantly. Artificially extending the study leads to panel fatigue and response dropout.<\/p>\n<p>If you are facing a question about the future of a market without hard historical data, <strong><a href=\"https:\/\/humes.pl\/en\/contact\/\">ask about an expert forecast using the Delphi method<\/a><\/strong> &#8211; Hume&#8217;s Institute specialists will help select the panel, design the rounds, and reach a credible consensus.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>You need to decide whether to enter a new segment, assess the pace of technology adoption, or estimate how a market will develop when it has no sales history or time series to model. Traditional quantitative forecasting has limited usefulness where there is no data to extrapolate, and it is precisely in such situations that [&hellip;]<\/p>\n","protected":false},"author":4,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"inline_featured_image":false,"footnotes":""},"categories":[912],"tags":[],"slowa_kluczowe":[],"class_list":["post-3599","post","type-post","status-publish","format-standard","hentry","category-badania-i-analizy"],"acf":[],"_wp_attached_file":null,"_wp_attachment_metadata":null,"wpml_media_processed":null,"_wpml_media_usage_in_posts":null,"_wp_attachment_context":null,"_oembed_35c905c64c03156f243b94f18c4eb80f":null,"_wp_attachment_image_alt":null,"rank_math_description":"The Delphi method turns expert consensus into a structured forecast for new markets and long horizons, using anonymized multi-round surveys.","rank_math_focus_keyword":"Delphi method","rank_math_contentai_score":null,"_wpml_post_translation_editor_native":null,"_menu_item_type":null,"_menu_item_menu_item_parent":null,"_menu_item_object_id":null,"_menu_item_object":null,"_menu_item_target":null,"_menu_item_classes":null,"_menu_item_xfn":null,"_menu_item_url":null,"_wp_page_template":null,"rank_math_og_content_image":null,"_wp_trash_meta_status":null,"_wp_trash_meta_time":null,"_wp_desired_post_slug":null,"rank_math_primary_category":null,"_acf_changed":null,"wp_pattern_sync_status":null,"_form":null,"_mail":null,"_mail_2":null,"_messages":null,"_additional_settings":null,"_locale":null,"_hash":null,"_config_validation":null,"_wp_old_slug":null,"rank_math_internal_links_processed":"1","_top_nav_excluded":null,"_cms_nav_minihome":null,"_thumbnail_id":null,"_last_translation_edit_mode":null,"_wpml_word_count":"1652","_dp_original":null,"_edit_last":null,"_edit_lock":null,"rank_math_seo_score":null,"_wpml_location_migration_done":null,"_wpml_media_duplicate":null,"_wpml_media_featured":null,"_wp_old_date":"2026-08-25","copied_media_ids":[],"referenced_media_ids":[],"rank_math_title":"Delphi method: forecasting without data | Hume's Institute","job_department":null,"_job_department":null,"job_location":null,"_job_location":null,"job_offer_external_link":null,"_job_offer_external_link":null,"footnotes":null,"inline_featured_image":null,"blog_podtytul":null,"_blog_podtytul":null,"blog_czas_czytania":null,"_blog_czas_czytania":null,"blog_dalsza_lektura":null,"_blog_dalsza_lektura":null,"slownik_krotka_definicja":null,"_slownik_krotka_definicja":null,"slownik_cytat":null,"_slownik_cytat":null,"slownik_na_stronie_glownej":"1","_slownik_na_stronie_glownej":null,"slownik_slowa_kluczowe":null,"_slownik_slowa_kluczowe":null,"slownik_w_praktyce":null,"_slownik_w_praktyce":null,"slownik_powiazane":null,"_slownik_powiazane":null,"slownik_kluczowe_punkty":null,"_slownik_kluczowe_punkty":null,"lang":"en","translations":{"en":3599},"pll_sync_post":{},"_links":{"self":[{"href":"https:\/\/humes.pl\/en\/wp-json\/wp\/v2\/posts\/3599","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/humes.pl\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/humes.pl\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/humes.pl\/en\/wp-json\/wp\/v2\/users\/4"}],"replies":[{"embeddable":true,"href":"https:\/\/humes.pl\/en\/wp-json\/wp\/v2\/comments?post=3599"}],"version-history":[{"count":1,"href":"https:\/\/humes.pl\/en\/wp-json\/wp\/v2\/posts\/3599\/revisions"}],"predecessor-version":[{"id":3600,"href":"https:\/\/humes.pl\/en\/wp-json\/wp\/v2\/posts\/3599\/revisions\/3600"}],"wp:attachment":[{"href":"https:\/\/humes.pl\/en\/wp-json\/wp\/v2\/media?parent=3599"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/humes.pl\/en\/wp-json\/wp\/v2\/categories?post=3599"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/humes.pl\/en\/wp-json\/wp\/v2\/tags?post=3599"},{"taxonomy":"slowa_kluczowe","embeddable":true,"href":"https:\/\/humes.pl\/en\/wp-json\/wp\/v2\/slowa_kluczowe?post=3599"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}