Abstract Background In the UK, about 14% of community-dwelling adults aged 65 and over are estimated to be at risk of malnutrition. Screening older adults in primary care and treating those at risk may help to reduce malnutrition risk, reduce the resulting need for healthcare use and improve quality of life. Interventions are needed to raise older adults’ risk awareness, offer relevant and meaningful strategies to address risk and support general practices to deliver treatment and support. Methods Using the Person-based Approach and input from Patient and Public Involvement representatives, we developed the ‘Eat well, feel well, stay well’ intervention. The intervention was optimised using qualitative data from think aloud and semi-structured process evaluation interviews with 23 and 18 older adults respectively. Positive and negative comments were extracted to inform rapid iterative modifications to support engagement with the intervention. Data were then analysed thematically and final adjustments made, to optimise the meaningfulness of the intervention for the target population. Results Participants’ comments were generally positive. This paper focuses predominantly on participants’ negative reactions, to illustrate the changes needed to ensure that intervention materials were optimally relevant and meaningful to older adults. Key factors that undermined engagement included: resistance to the recommended nutritional intake among those with reduced appetite or eating difficulties, particularly frequent eating and high energy options; reluctance to gain weight; and a perception that advice did not align with participants’ specific personal preferences and eating difficulties. We addressed these issues by adjusting the communication of eating goals to be more closely aligned with older adults’ beliefs about good nutrition, and acceptable and feasible eating patterns. We also adjusted the suggested tips and strategies to fit better with older adults’ everyday activities, values and beliefs. Conclusions Using iterative qualitative methods facilitated the identification of key behavioural and contextual elements that supported engagement, and issues that undermined older adults’ engagement with intervention content. This informed crucial revisions to the intervention content that enabled us to maximise the meaningfulness, relevance and feasibility of the key messages and suggested strategies to address malnutrition risk, and therefore optimise engagement with the intervention and the behavioural advice it provided.